FigS14.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "456ac77a-57f9-4fe4-b81e-691b5af4107e",
   "metadata": {},
   "source": [
    "# Path to Zenodo files folder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0c91ad4e-8ff7-42d0-a940-47bd0ef5b8e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "path = 'PATH/TO/ZENODO/FILES/FOLDER'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e77ea23b-1dc8-49c4-ae83-02233aede633",
   "metadata": {},
   "source": [
    "# Figure S14"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01e1bfd9-f7c6-4e9e-9f5d-7617dc7eb123",
   "metadata": {},
   "source": [
    "## Panel a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "625fb29b-5d74-42c8-9a94-820537ea23a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "predicted_phenotypes = pd.read_csv(path+'Fig7'+'/FYN_coreDTS_bPCA_order2_sparseregularized/task_4/predictions/predicted_phenotypes_all.txt',\n",
    "                                  sep='\\t', lineterminator='\\n', low_memory=False)\n",
    "\n",
    "predicted_abundance = predicted_phenotypes.loc[predicted_phenotypes['Abundance']==1]\n",
    "predicted_binding = predicted_phenotypes.loc[predicted_phenotypes['Binding']==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "92fefc9b-c442-42dd-9f80-1d11c087f318",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalize data\n",
    "\n",
    "from seaborn import kdeplot\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "RT=303*0.001987\n",
    "\n",
    "predicted_abundance_scaled = predicted_abundance.copy()\n",
    "predicted_binding_scaled = predicted_binding.copy()\n",
    "\n",
    "def normalize(values, bounds):\n",
    "    return [bounds['desired']['lower'] + (x - bounds['actual']['lower']) * (bounds['desired']['upper'] - bounds['desired']['lower']) / (bounds['actual']['upper'] - bounds['actual']['lower']) for x in values]\n",
    "\n",
    "# Normalize abundance fitness measurements\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_abundance, x='fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "\n",
    "predicted_abundance_scaled['fitness']=        normalize(list(predicted_abundance_scaled['fitness']),{'actual': {'lower': mode, 'upper':         0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_abundance_scaled['fitness1_uncorr']=normalize(list(predicted_abundance_scaled['fitness1_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_abundance_scaled['fitness2_uncorr']=normalize(list(predicted_abundance_scaled['fitness2_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Normalize abundance fitness predictions\n",
    "\n",
    "for index, row in predicted_abundance_scaled.loc[predicted_abundance_scaled['WT']!=True].iterrows():\n",
    "    predicted_abundance_scaled.at[index,'predicted_fitness']=row['fold_'+str(int(row['Fold']))]\n",
    "    predicted_abundance_scaled.at[index,'predicted_ddG']=row['fold_'+str(int(row['Fold']))+'_additive_trait0']*RT\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_abundance_scaled, x='predicted_fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "predicted_abundance_scaled['predicted_fitness']=           normalize(list(predicted_abundance_scaled['predicted_fitness']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Obtain mean inferred folding ddGs\n",
    "\n",
    "predicted_abundance_scaled['mean_additive_trait']=predicted_abundance_scaled[['fold_'+str(i)+'_additive_trait0' for i in range(1,11,1)]].mean(axis=1)\n",
    "predicted_abundance_scaled['std_additive_trait']=predicted_abundance_scaled[['fold_'+str(i)+'_additive_trait0' for i in range(1,11,1)]].std(axis=1)\n",
    "\n",
    "predicted_abundance_scaled['mean_ddG']= predicted_abundance_scaled['mean_additive_trait']*RT\n",
    "predicted_abundance_scaled['std_ddG']=  predicted_abundance_scaled['std_additive_trait']*RT\n",
    "\n",
    "# Normalize binding fitness measurements\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_binding, x='fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "\n",
    "predicted_binding_scaled['fitness']=        normalize(list(predicted_binding_scaled['fitness']),{'actual': {'lower': mode, 'upper':         0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_binding_scaled['fitness1_uncorr']=normalize(list(predicted_binding_scaled['fitness1_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_binding_scaled['fitness2_uncorr']=normalize(list(predicted_binding_scaled['fitness2_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_binding_scaled['fitness3_uncorr']=normalize(list(predicted_binding_scaled['fitness3_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Normalize binding fitness predictions\n",
    "\n",
    "for index, row in predicted_binding_scaled.loc[predicted_binding_scaled['WT']!=True].iterrows():\n",
    "    predicted_binding_scaled.at[index,'predicted_fitness']=row['fold_'+str(int(row['Fold']))]\n",
    "    predicted_binding_scaled.at[index,'predicted_ddG']=row['fold_'+str(int(row['Fold']))+'_additive_trait0']*RT\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_binding_scaled, x='predicted_fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "predicted_binding_scaled['predicted_fitness']=           normalize(list(predicted_binding_scaled['predicted_fitness']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Obtain mean inferred binding ddGs\n",
    "\n",
    "predicted_binding_scaled['mean_additive_trait']=predicted_binding_scaled[['fold_'+str(i)+'_additive_trait1' for i in range(1,11,1)]].mean(axis=1)\n",
    "predicted_binding_scaled['std_additive_trait']=predicted_binding_scaled[['fold_'+str(i)+'_additive_trait1' for i in range(1,11,1)]].std(axis=1)\n",
    "\n",
    "predicted_binding_scaled['mean_ddG']= predicted_binding_scaled['mean_additive_trait']*RT\n",
    "predicted_binding_scaled['std_ddG']=  predicted_binding_scaled['std_additive_trait']*RT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "6062698f-a9f9-4360-a34e-45a52882dcb9",
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_abundance_scaled = predicted_abundance_scaled.rename(columns={'predicted_fitness': 'predicted_stability_fitness'})\n",
    "predicted_binding_scaled = predicted_binding_scaled.rename(columns={'predicted_fitness': 'predicted_binding_fitness'})\n",
    "predicted_abundance_scaled = predicted_abundance_scaled.rename(columns={'fitness': 'stability_fitness'})\n",
    "predicted_binding_scaled = predicted_binding_scaled.rename(columns={'fitness': 'binding_fitness'})\n",
    "predicted_abundance_scaled = predicted_abundance_scaled.rename(columns={'mean_ddG': 'folding_ddG'})\n",
    "predicted_binding_scaled = predicted_binding_scaled.rename(columns={'mean_ddG': 'binding_ddG'})\n",
    "predicted_stability_vs_binding = predicted_binding_scaled.merge(predicted_abundance_scaled, left_on='aa_seq', right_on='aa_seq')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1cc2e797-1e94-48a5-bd3f-8f485441d87c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib\n",
    "#matplotlib.use('QT4Agg')\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import axes3d\n",
    "\n",
    "class MyAxes3D(axes3d.Axes3D):\n",
    "\n",
    "    def __init__(self, baseObject, sides_to_draw):\n",
    "        self.__class__ = type(baseObject.__class__.__name__,\n",
    "                              (self.__class__, baseObject.__class__),\n",
    "                              {})\n",
    "        self.__dict__ = baseObject.__dict__\n",
    "        self.sides_to_draw = list(sides_to_draw)\n",
    "        self.mouse_init()\n",
    "\n",
    "    def set_some_features_visibility(self, visible):\n",
    "        for t in self.zaxis.get_ticklines() + self.zaxis.get_ticklabels():\n",
    "            t.set_visible(visible)\n",
    "        self.zaxis.line.set_visible(visible)\n",
    "        self.zaxis.pane.set_visible(visible)\n",
    "        self.zaxis.label.set_visible(visible)\n",
    "        \n",
    "\n",
    "    def draw(self, renderer):\n",
    "        # set visibility of some features False \n",
    "        self.set_some_features_visibility(True)\n",
    "        # draw the axes\n",
    "        super(MyAxes3D, self).draw(renderer)\n",
    "        # set visibility of some features True. \n",
    "        # This could be adapted to set your features to desired visibility, \n",
    "        # e.g. storing the previous values and restoring the values\n",
    "        self.set_some_features_visibility(True)\n",
    "\n",
    "        zaxis = self.zaxis\n",
    "        draw_grid_old = zaxis.axes._draw_grid\n",
    "        # disable draw grid\n",
    "        #zaxis.axes._draw_grid = False\n",
    "\n",
    "        tmp_planes = zaxis._PLANES\n",
    "\n",
    "        if 'l' in self.sides_to_draw :\n",
    "            # draw zaxis on the left side\n",
    "            zaxis._PLANES = (tmp_planes[2], tmp_planes[3],\n",
    "                             tmp_planes[0], tmp_planes[1],\n",
    "                             tmp_planes[4], tmp_planes[5])\n",
    "            zaxis.draw(renderer)\n",
    "        if 'r' in self.sides_to_draw :\n",
    "            # draw zaxis on the right side\n",
    "            zaxis._PLANES = (tmp_planes[3], tmp_planes[2], \n",
    "                             tmp_planes[1], tmp_planes[0], \n",
    "                             tmp_planes[4], tmp_planes[5])\n",
    "            zaxis.draw(renderer)\n",
    "\n",
    "        zaxis._PLANES = tmp_planes\n",
    "\n",
    "        # disable draw grid\n",
    "        #zaxis.axes._draw_grid = draw_grid_old"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "bfda15c0-20c7-4ed7-b008-ca2b9757459e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAANwAAADcCAYAAAAbWs+BAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAACJcUlEQVR4nOyddZiVZfrHP6cmz3R3J8MM3TAgoCIS1qKwJrIYWCu28ltdY3VXzLXXWItFLEJF6RoaBpju7jwTp5/fH8+cAyOggEO4O9/rmmtmzhvnPfF97/v+3vEohBCCfvSjH+cEyvN9Af3ox/8S+gnXj36cQ/QTrh/9OIfoJ1w/+nEO0U+4fvTjHKKfcP3oxzlEP+H60Y9ziH7C9aMf5xD9hOtHP84h+gnXj36cQ/QTrh/9OIfoJ1w/+nEO0U+4fvTjHKKfcP3oxzlEP+H60Y9ziH7C9aMf5xD9hOtHP84h+gnXj36cQ/QTrh/9OIfoJ1w/+nEO0U+4fvTjHKKfcP3oxzlEP+H60Y9ziH7C9aMf5xD9hOtHP84h+gnXj36cQ/QTrh/9OIfoJ1w/+nEO0U+4fvTjHKKfcP3oxzlEP+H60Y9ziH7C9aMf5xD9hOtHP84h+gnXj36cQ/QTrh/9OIfoJ1w/+nEO0U+43xmsVitCiPN9Gf04Q6jP9wX049TQ1dVFaWkplZWVODo6Eh4eTkBAAC4uLuf70vpxGlCI/tvlBY3u7m7Kysqora3FwcGBjo4OPD09MRqNmM1m3N3dCQgIwN/fH41Gc74vtx+/gn7CXaDQ6/V2omk0GrtF2759O8nJyfj4+NDU1ERdXR3Nzc0A+Pj4EBAQgI+PD0plf7RwIaLfpbzAYDAYKC8vp7q6GrVaTXR0NMHBwahUKiwWCwBCCFQqFf7+/vj7+2M0Gqmvr6euro6srCzUajX+/v4EBATg7u6OQqE4z6+qHzb0E+4CgdFopKKigqqqKpRKJZGRkYSEhKBW//pH5ODgQGhoKKGhoXR2dlJXV0ddXR3V1dU4OzsTEBBAQEAAzs7O5+CV9OOX0E+48wyTyWQnGkBYWBhhYWGnRLQTwdXVlejoaKKiomhtbaWuro6KigpKS0vx8PCwx3tnev5+/Db0v+vnCWazmcrKSiorK7FarYSGhhIWFvaLwsfpuIYKhQIvLy+8vLyIi4ujsbGR2tpa8vPzKSwsxMfHh8DAQLy8vPrjvXOIfsKdY1gsFqqqqigvL8dqtRIcHEx4eDgODg6nfI7T1blUKpXdrTQYDNTX11NbW8vhw4fRaDT4+/sTGBiIVqvtj/fOMvoJd45gsViorq6mvLwcs9lMUFAQERERODo6ntPrcHR0tLutHR0d1NbWUl9fT1VVFS4uLgQGBuLv74+Tk9M5va7/FfQT7izDarVSU1NDWVkZJpOJwMBAIiIiLogvtFarJTY2lpiYGFpaWqitraW0tJTi4mI8PT0JDAzE19e3P97rQ/S/k2cJVquVuro6SktLMRgMBAQEEBkZ+ZuUwrPl7ikUCry9vfH29sZsNtPQ0EBdXR25ubkolUr8/PwICAjAy8ur3+X8jegnXB9DCEFdXR1lZWV0d3fj5+dHZGQkrq6uffocZwtqtZqgoCCCgoLQ6/X2FENdXR0ODg72WFCr1Z61a/hvRj/h+ghCCBoaGigtLaWrqwtfX18GDBjwu/5iOjk5ERERQXh4ODqdjrq6Ompra6moqECr1drJdzqCz/86+gn3GyGEoKmpiZKSEjo7O/H29iYpKQk3N7fzfWl9BoVCgbu7O+7u7sTExNDc3ExdXR3FxcUUFRXh7e1NQEAAvr6+qFSq8325FzT6CXeGEELQ3NxMaWkpOp0OT09PBg8ejIeHx1l7zgshflIqlfj6+uLr64vJZLLHezk5OahUKnu85+npeUFc74WGfsKdAVpaWigpKaG9vR0PDw/S0tLw8vI6Z89/odSbazQagoODCQ4Opru72+5y1tbW4ujoSEBAAIGBgf0tRMegn3Cngba2NkpKSmhtbcXNzY3U1NR+5a4Hzs7OREZGEhERQXt7u72Ws7y8HDc3N3tJ2f96vNdPuFNAe3s7paWlNDc3o9VqSUlJwcfHp59oJ4BCocDDwwMPDw9iY2NpamqitraWoqIie7wXGBj4P9tC1E+4X0BHRwclJSU0NTXh4uLCgAED8PX1Pa9E+z2R3JbD8/Pzw2g00tDQQG1t7f90C1E/4U6Azs5OSktLaWhowNnZmaSkJPz9/S+YL8WFEsOdDhwcHAgJCSEkJISuri5qa2v/J1uI+gl3DLq6uigrK6Ourg4nJycSEhIIDAy8YIj23wIXFxd7C1FbW5s9t3dsC5Gfn99/5ciI/hELyHEGpaWl9mqKiIgIAgMDL8gYY/PmzcTGxhISEnK+L6VPYbFYaGxstI+MUCgU+Pr6EhAQgLe39wX5WZwJ/qctnMFgoKysjJqaGjQaDTExMQQHB1/QH+5/q7U9UQtRXV0dR44csbcQBQQE4Obm9rt+D/4nCWc0Gu1zQ1QqFVFRUYSEhPxuqiT+252Sn7cQ2Wo5bS1ENmJeCB0Xp4v/KcKZTCbKy8vtc0MiIiJOeW5IP84PtFotWq2W6OhoWlpa7IXhJSUleHp62uO938tn+D8Rw5nNZioqKqisrASwD9z5PQblW7ZsITo6mtDQ0PN9KecNZrPZHu+1tLTYy81sIyMuZJfz93FbOEOYzWaqqqqoqKjAarUSEhJCeHj475Jo/TgKtVpNYGAggYGBGAwGe0nZoUOHcHBw6DUy4kLDfyXhbHNDKioqMJvNBAcHExER8V9RVqRQKP7rY7jTgW3s+8/jvcrKSlxdXe0jI871KIuT4b+KcFarlerqasrKys7r3JB+nHsoFArc3Nxwc3Ozx3u1tbWUlJRQVFSEl5eXfWTE+RTH/isIZ5sbUl5ejtFoJCAggIiIiP/6qoV+nBhKpRIfHx98fHzsIyNqa2vtLUS2eO98tBD9rgknhKC2tpaysjL0ej3+/v5ERkb2t4P0w45jR0bYWohsP46Ojuj1etzd3Rk6dGifPu9//vMfnJ2dmTlzJtu3b2f16tWMHTv296lSCiGor6+ntLT0rM0NuVCxbds2wsPDCQ8PP9+X8ruFEAKdTkdtbS3XXnstDQ0NVFRU9Mm5u7u77bWhISEh7N+/Hy8vL1pbW3FwcPh9Wbifzw3x8fH53c8N6ce5x7EjI0wm01lJoOt0OgYOHMiDDz6ISqVCCCFHzPf5M50F2OaGlJaW0tHRgZeXF4mJibi7u5/vS+vH7xzd3d196hnZBBkPDw+2bdtGTU0NSUlJgGz3unCLBjk6N2T//v0cOXIElUqFs7Mz8fHx/WTrx29GeXk5HR0dfUo4W+pp8+bNtLe34+TkxCeffALIhVYuWMK1tLRw8OBBDh06hEKhIC0tDR8fH7q7u9m/f/+JD+rqgqIiOHQI8vKgoED+7NkDhYXyJycHKirg9xe6Av15uL7EkiVLqKurY//+/Tz55JNYrVb7tsDAQNRqNa6urvYFL0FaKXd3d9RqNX5+fic994EDB2hoaLBPOPv3v/9NbW0tiAsMra2t4sCBA2Ljxo1i7969orGxUVitViGEEBaLRRw4cEDodDohmpqEyM8/+vPjj0KUlgphscifAweEMBrlSVtahCgsFD0nEWLnTvl/QYEQGRnyt23fCxzbtm0TpaWl5/sy/mvg7OwsvLy8REBAgP2xZ599Vri6ugohhIiJiRFTpkyxb5s9e7YIDQ0VQgjh4eEhHnjggV7nM5vN9vP+9a9/FWazWWg0GgEIHx8fcUFZuKVLl/LUU09hMplISUlhyJAhR2eH6PUoCwsZ5OqKtqYG6uvBYoHYWIiLgylToLMTOjpAqYRBg6SV6+4GT0/w9obS0qPbjEZ57IAB8rGKCjhyBLKyfrfWrx9nhiuuuEJanx6sWLGCYcOGAXDjjTf28qh27tzJlVdeCUB6ejqrVq064Tm7u7u59957mT59Oh4eHgghZN3nWXwdp42amhp0Oh3Dhg2Ts0P0esjPlz+VlZIkPj6SYImJEBMjSdLVBQoFJCdDYyM0NMgTpqRASYnc7uUFWq0klqMjBAdLArq5gbMzuLvL/Q0GSdS8PEnofvxXw2w24+np2eux9vZ2fHx8AIiIiECv19u3dXd3ExwcDIC/vz86na7XsTbRRKPRcNVVV7F+/XquueYaQGoSF5RK6ebmRmNj49Hsf06OJEZg4NGdqqpkXBYXBxoNDBwIBQXogQO7dmHQ63FobsbT25vSri6cXVwYlpWFW1ISODlBbS3U1MDgwZKI9fUQFAS5ueDhIS1eWZm0fhs2QHi4fK4LpAK9P4brOwghsFgsx80U9fDwoKmpCYCSkpJeaQMXFxeqq6sBqK2tPemE7TvuuIN//vOfeHt788Ybb7BlyxY0Gs2vE66oqIimpiba2tpoa2vDaDRSW1tLa2srBoOB5557rk86pI1GI1qttvcdY8gQaGqSbl5cHDg4QEgItLfDmjWSDE5OWCwWuisraW1vp0GrxSE8nK/WrEGr0TD0iiswjxsniRoeLolUVCTJ7OIiLWBgICQkSKuWkgKurtDaCunp0rrm5EgLGRT0m19nPy4cWCwWhBB2a2bDVVddxTPPPAPAxx9/bHcvAUaOHMlXX33FK6+8wtatW1m4cOEJz/3yyy9z11130d3dDcDw4cMxGAwnF01sQoVGoxGxsbFixIgRYsqUKUKhUIgrrrhC3HfffUKhUIjm5ubfHrkKIa6//noRFxcn0tLShPHnAobVKkRenhC7dkmBpLRUPpafL5qys8WKFSvEvn37hL62VhiPHBGffPKJmDdvnnh1yRIhSkpER0eHKCwsFOZ9+4ToCWpFdrYQer38OzNTiij79wuxdat87PBh+Rzl5UK0tQlRUSHExo1HjzlP2L59uygpKTmv1/DfAr1eLwDxxRdfHLctICBAqFQq4eLiIpqamoSnp6cQQgidTie0Wq1QqVTCx8fnhOc1GAxi9uzZwsXFRahUKiGEEP/4xz/E4MGDxa8SbsKECb0eP1axufjii0V1dfVpvswTo6mpScydO1c4OjqK995778Q7NTcLceTIUdIIIbpKSsT2f/9bFBcXCyGE0Le0iB9ffFEsffFF8fXXX4vMDRvEX26+WYwdO1bMve46se7FF0VhQYE8ODNTkspoFCInRz6WmyvEnj3yeWyPZWUd/Z2XJ0RlZZ+85jNBP+H6DnV1dQIQmzdv7tPzZmRkCKVSKe6//36hVCqFEEJs3rxZKBSKk6uUoidOMJlMHDp0yP54Xl4eq1atslfot7e3H3fs4sWLGT9+PPPmzcNoNNofN5vN3HTTTYwfP5577rmn1zHe3t5cd911xMfHc/3115/4ory8pDCSny9dTcA5MpIxV1xBVEkJFBTg2NjI1GnTuNHHhwijEVNDA3WHDuGydy/lGzfyz/feY9mtt1KYl0e+SkXRmjUyFvTxkfFdQoJ0NRMToa4ODh6UrmhZGcTHSwXT2Rm2bTsvamZ/DNd3qKurAyCoj0OFiooKlEolzz77rF2PKC4uRqFQ/LpKmZSUxGeffcbKlSv529/+xrBhw/j++++59957mTRpEt7e3r32P3DgALW1tWzdupXk5GRWrFhh37Zq1SpCQ0PZunUrXV1d7Nixo9exWq2Wjo6OX24UVSggKUkqltu2yWR2bS0MHy5TAGFhkJSE9/XXMzg+niFXXUXQrFnoIyLosFqp1GqpDQvj4KuvUrhuHVX5+bB9O/j6ytjQYJDqZ0mJjOGsVqlsVlXJ53dwkOJLSIhMqJ/ghtOP3wdsqYDAY0W5PoCPjw++vr6MGDECq9XKypUrefLJJ3Fzczs54WxCyKuvvorZbGbJkiVkZmbywQcfcP311xMVFcWjjz56XLY9IyODiy++GIBLL720F6l+aRtIlbKjo+PUXlVQkMyndXdDdLSU91NTpSCi08ncWloaeV98QYCfH16JiQwdPJi7rr2Wq+bP54gQbNm/n/i5c6VocvCg7SJl2sDREdrapKVzdJRiy3ffSWtXWgpRUfI5m5uPkrEfvyvYLFxfNyhfdNFFPPXUU+Tn56NQKLjiiivo6Oigvr7+11VKjUbDP/7xD/v/DQ0NREVFMXr0aFpbW2lvb+9V19ja2mrPU3h4ePQqi2ltbbXv+/NtwPEq5a9Bq5Wq4pEjknSurtLlLCwEsxm8vHAdMwbF888THBSEQqnE39mZD199lcySElzr62m4+25GTZ3K9IgIggcNku7ihg3SylVWyjSB1SrzdMOGSavm5iZzfT4+kti1tfKnj3uq+nF2YUtBnY0O8AULFrBgwQJycnIIDg5Gq9XS1tZ2csIJIVAoFKxfv54dO3ag0WgwGAwolUpqamp44IEHKCwsJCAgoJds6uXlZY/rWltbe7mcv7QNpIXT6/WYzeZTH3umUMhcXGGhzKvZmk8PHgStljBvb265806i3nkHZXIyGzdswHL4MI06HVVWK4GZmbyZn0/g//0fqi1bcEhNxSsqSpK3uVlavNGjZWpiwABp9Xx8YPduuOwymb9LSZHPd+iQdHfP4pCi/hiu79Dc3IxSqezzwb8HDx7kz3/+M1lZWYD0Fo1GI0OGDDk1C+fm5oafnx+Ojo64uLhgtVrta339vJdo1KhRvPjii9xwww2sXbuWsWPH9tr2448/MmHCBNauXcstt9zS61hbX1tHR8dx2f9fRWystEhKpUyWx8baK0lUPj6MePxxiteuJe3qq+l0cGDMihVkGgzkt7URqtHwzvvv82lXF8GBgbz4yScosrNlcjwzEw4ckEnx+noIDcXa3Ixy7Fj48UfpxlZWypjOapWCjrd3f87ud4Dm5mbUanWfjVkwGo04ODhw//33s2nTJsaPH4+npycGg4GWlhYGDBjw68XLRqNRNDU1iZqaGlFcXCwKCgrE/v37RXl5+UmPuf/++8W4cePE3LlzhcFgEH/605+EEEKYTCZxww03iHHjxom77rrruOPMZrMAREVFxZmosRK1tUKUlR39v7hYFjoLIURHhxAFBaKrq0tcPHGiGOriItRqtUh0dBTTXF1FolIpxjg7iz/dcovI37dP5vuEkOmA8nIh1qwR+/buFWtfflk0NzXJtEFBgRA2WTk7W/7+6aezljrIyMgQRUVFZ+Xc/2v44x//KFxcXPrsfAaDQQghxOTJk0VKSsoJ9znpiAWr1YpSqWTlypW88847+Pn5oVAoaG1tpbKykgcffJCrr74ai8XSpz6wq6sr+/btIzEx8cxP0tQkXb/oaPl/UZFUIT085LauLlq0WjZ/8w0v/N//saemhhiVimqjEU8hmBYczF2ff05KcLBUPTs7pSrq6kr1F1+Q5+HB6Lg4nAYMkK6sq6usUhk+XD6vg4P8DTKV0IfYuXMn/v7+RNteWz/OGLNmzWLz5s20trb2yflsFm7KlCls3ryZkSNHMnXqVHx8fPDz8yM9PR31F198IYsq1Wri4+NJSUkBjqqUI0eOxNfXFxcXFxwdHWlsbGTNmjX2i+zrqUdubm6nJ5ycCD4+MrbLyJBEUyhg585edZleVivps2fz6RtvUNLURKubG9GNjWRaLKzv6uLemhq+3LKF5LAwkm68EbKzISCA4PHjCVappBtpMoFaDX5+smi6qEj+P3CgjP+iomD9ehg/XpKwD9Afw/UddDpdnyqUNsPj6emJ1Wplx44d7Nixw/55DRs2DPXXX39Nd3c3Li4uxMXF8eOPP7Jo0SJ7Lsy2cMKxKC4uZs2aNdx66619/uHbcnG/Gd7eUtJvaZGWLjZWqpkqFej1kJ2NWaNh6ODBRHh50RgUxKG9e4nJzqaotZVHliyhwcuL6VFRJEVFYRg0iPxvviF+2jQci4tlnefatTB9OnVbt+I+ciTOZWXy3AcOSLI1NUnV9OBBeQ2+vr/9dfWjz3C2CLdw4cJe+edjoX777bftFc/Z2dmsXLkSk8lkJ5xOp2PPnj0YDAba29tpaGhgz549TJgwAaDPFZ4+sXA2eHlJIaOsDCIipMp4+LAUOuLiOPjii1SpVLi4uDC4tpaIkSPZajbTWVbGtsZG4uvqWKNScVVjIzuffRZvlQpCQhgYGirTDiNG0LlpE9l5eWi6uxmXlASRkbBjB/j7y6R4UpIshDYapZVMTu6b19aP34yurq6zMrt0zpw5XHTRRb1I5+vry8SJE1Hff//9ODs7c+WVV5Kenk5yzxdC9KQF8vLyuPXWW0lKSkKr1eLp6cnUqVP5wx/+APS9S9lnFs4GHx/Z11ZVJZXExERZ/Z+UxJC5cyl+/nlWmc0UG40kmUwMiYsjpr2dZe3tFHV14XjgAO9u2UKExcLawkJuam9n4EsvyRahpCRcWlrwNhgI7eyUVi07GyZNki6sm9vRbgS9Xub4duyQhO+fNHbeYZtJ2Vd4/fXXWb16Na2trWzfvp277roLk8nEqFGjaG9vx9nZGcXevXvFpk2b+O6773jjjTdISEiwk+1YWCwWqqursVgsBAUF4ejoeML9fisuu+wyZs2addK2hzNGVZXMj/n7S4vT1gYRERTu3Mme7dvZkJvLEK0WZUwMyz//HO/Dh8nU6VADZS4uDPbyokur5ZIxY7goIoLJc+ag9PCQZV56vcwBVlXJhLkth+flBVu3wsUXQ3GxtG45OZJ4ZrN0c08Tu3fvxtvbm9gzOLYfvREcHEx0dDTbtm3rk/O9/PLLvPDCC9TW1qJSqVAoFPYWoKioKL7++mvUQ4cOZejQofz000+UlZWRkJBw3IkyMjJ477336OjoQK1W4+3tzTXXXGN3K/sSfW7hbAgJgb17JSm0WlkZUl+Pd2Qk0UVFjJo2DaNSifnQIfanpLAlM5MGIAwY1NXF9q4uhrq48MWWLbSGhhIyYQLJJSUwbZrdYtLVJeO2+nqZw8vJgcsvl7HewIFSVAkPlwKLh4cUVFJS4Gcxcj/ODWw9mH2FK6+8knvvvZfFixfz6KOPHlfYAaDMzMxkxYoV9uT2z9HQ0MDdd9/NxIkTee211/jb3/5GQkICjzzyCCC7CfoSfRrD/RzDhkm1MDJSKoeurtTs3IlRr8dp+3YSYmOJmTCBcKWSFq2WCI2Gw4AOGATUdXWhKyriq7178dVqaU1Kgu+/l0nu0lKIjORAWRn7du2SlSmenkfjOKNREs7BQdZ/enuDbY237GwZZ/bjnMJoNJ60Y/tMcMMNN1BSUsLhw4cZMWIEycnJDB48mLFjxzJ27FhKSkpQv/TSS1RVVXH33XczePBgoHdc5uDggKenZ6+WmQULFrB161aAPl9r7axZOBuSk48KJ0lJJJvNtEyYgLejI2zZQp3ZzOioKA6OGkVZQQEJxcUc1utJRC7EEA9s6Ozk6ssuo9HXl2dvvJGLAwJw0eshIoKuujoOC0FqbS0aZ2fpviYlSWuXni4LoCdMkBUpiYmSbAMGyI4FvV6qqNHRsmKmH2cVJpPp9CuafgFFRUVotVoKCgqwWq2YTCZaW1uxWCzo9Xr0ej3qBx54gJiYGBQKBeXl5XR3dxMQEGC/EK1WS2BgIEuWLGHEiBE0Njby7bff0tjYyF/+8hc6Ojp6FTf/Vri5udFgGwJ0NqBQyC96Xh4kJKCIicG7tlbGXrGxqDs6WPWvfzGypYVhs2ax5oMPKKqtpQFQAQeBqcDWxkZi29p4Y906Dh06xKQZMxjX3Y3fqFEMzsxkdXU1E11c8LJYZM7O21t2MYwYId3alhYZ5/n7y767QYOkKxoWBj/8cDS+8/Do5XL25+H6DieaZ/JbUFdXh5+fHxqNhry8vBPuo16zZg3l5eX29dRqamqYM2cOkyZNAmRuoaysjN27d/PTTz+h1WpxcXFh9OjR6HS6Pp+ArNVqKSkp6dNzHgcHB5kTsymXjo6SAOHheO7aRfCwYbgaDNw8dSqK6mo0q1ezoamJJCAHKAMigGBnZ9RAU0cHP+zdS/L06UR1dqKzWjGp1eTl5jLqxhvl/JUxYySxkpPlc40ZI61daqqsZPHzk9atq0tawLo6eRPIyLA329LThdGP3w6r1drnhANITEwkPz+fyZMnExAQgI+PD8HBwfj7+zN//nzUTk5OKBQKCgsL8fLy4vLLL7e319iwZcuWPr2oX8JZjeGOhY+PrO6vq5OqYl0dDBmC66BBjC8ooBD4z1df0ejnR2pcHFaTiYz2dqKAfCABONDezsiNG/nA1ZVZ1dX8TaHgnqlTSUpIwDk/n+qgILJXraLNzY1R+fkoXFygvFzmBCsrYepUmV5QqaSAMnWqdHdTUmTXQ0OD7FTIzJTELChA2dUlY8N+/CZYLBasVusvTk8+XTzxxBMsXboUkKsc2RRKm0cyf/581HfffTcAy5cvZ+/evVx++eXHnWjZsmU0NTVhMpno7u7GYDDQ3NzMP/7xD15//XUWLFjQZ8HnWY/hjkVqqqw+iY+X+bLvvoPBg0lMS+PQjz/yQ26uHNMwdCixBgOaujraGhspNxrpArqAI1YrA3Q6yuvqqN+2jeu/+QY8PZk3YAABgwfzyfbtqGJiGHHDDagiI2Wv3ejRsrNcpZK5Oj8/+XvbNul67toFo0ZJMrq6SgL2xJ0OBw+i6OO78v8iLD0zR/uacE888QSzZs3i22+/PeE+yvz8fPbu3cv69evtEqnZbAaOzjXJzc2lqKiI9vZ21Go1/v7+JCUloVKpGDp06Kn3rp0CzpmFs8EWz0VH28c0uAhBXHs7s6+6iul/+AO+Dg7UeXvj5O1Nt6Mj6UAFMmVQCrQCtZWVmA8dYmdVFcbiYrYIgbW2lujQUK5KS0OVlCSVyOnTJXnMZvm8kZFS4QwLk8SLi5MWzJYb2rABqqtlTJeTgyE6GnVVlXRD+3HGaGlpAeQw176C0WjEYrHw7bff0tbWRnd3N93d3fbHAdRvvvkm1dXVBAcH86c//Uk+2EMgm1q5ZMkSTCYTJpPJPhSovb0dk8lEenp6n10wnGMLB7LY2MdHupQBAdDYiENKCsNdXRlfWIgiNZWsnTuZMnMmjdu3k6XTYdTpGAmUAz5APVJQUXd3MwDQREUxMDiYxEsvJUSvZ5BaLV1INzcObt2KCxAfGCgJt3u3VDFLSqSlPXJE5uxKS6W1i4mRXeYmExiNuGVkIGzqZmKitIz9OG3Y5pn05QAhWzmkxWLBw8PjhPuo//CHP+Dq6kpUVBRubm69qkdsf7/++uvk5OTg5OSEEILy8nIsFguvvfYa4eHhfVpxcs4tHEjLkpsriddj8bRJSWC10trcjLm1lZaMDLSDBpG8bx8bXV3RdHfjZrXiA2QgY7psYAjgrNPRlpfHpw4OuFVUoJ8/n3F5eRATQ+m2bWQJwc1mM8FTpsjC5spKOQ3a21tavJISWSZ25IgkY1qaJGBaGp0tLTgYDNIaFxZKIp4gwdqPX8aNN94IcFZWzf2ldjXlwIED+eqrr3j22Wf58ccfTyg7p6amkp6ezkUXXcRll13GpZdeikajsY+D7kuccwtnQ0KCzJUplbIkq6mJFl9f9qxbR31kJKacHPJ++onajg48lEqMXl50qNXUA6OBSiAUKajkVFez/ccf2ffxxxyoq6N81y62dHayf+9eRFsbSWo17V5elG/dSo5aLVMVEyfKMjDbQiVNTTI/d+SILB8LDobCQqxaLRY/P2kdY2NlquGYhSj68euwKZQgJ3Y9/fTT9m3Dhg074TJVAIsWLUKhUKBWq0+af96yZQsJCQkkJyczZcoUpk+fzrx583j77bepqqpCvX79er799ltuv/12nnrqKS6++GK7xbJZrYkTJx534m3btpGTk8PgwYPPioU7G3WavwiFQlq4LVsgNJTOxkY2V1fT1NVFWkgImVOmkFRZiY+DAy4ODjyzZg1xVit5gBsQ1XOaSiDUamVTVxeTSktRaLVkfP89lvh4Rnt44D1+PJOQM+q3HDxIm4cHSePGyZzc1KkyzktLg40bpZqpUslqlvh4LGYz6rw8SrVaQuPjcVq7VrqcNTWyNvQEZXn9OB5KpZL58+dz//3389RTT3HRRRcB8Pnnn9PU1ITZbCY2NpY5c+bw008/9To2ICCg10o7NtgasdeuXUt+fj5KpZKysjL0ej1Wq5XPP/+cRYsWoV67di3PPvss6enpvPrqq8DxHQCVlZXk5+fT2dlJW1sbTU1NNDU12f3fviSGVqvFarWi1+vPSuvELyI4WFqMwEBa8/Pxy8qixtub4K4uEpKTaTebGTB+PNU//cRAq5UCYICbG9k6HSGAFulSHgZigO3AmKoqqnU6Gg4fRjdxIncPHIjLnDmQm8uggQOxlpTI2DE7WxLe3Z3m0lL2KBRMMRhQDR6MCA2lOT+f4tZWOlta0B05wjc1NQRFRJBSUYHPxInyuo8ckVbxAll45EJGU1MTarWaRx991P79/fbbb5k6dSogXc6XX375uOPq6upQq9UkJyf3GpBsQ0lJCcHBwVQdMzpxzpw55OTkEBMTg7q4uJhJkybR0dFhF0BsF2CbnvXRRx/x1ltv2dfVdnBw4KabbmLcuHG99u8L2NILOp3u3BMOpHCRk0PIpZfi4+9PZGUl67OycGtoIO6qq2jftYufXFwI8fbG2WSiW6/HCGiAI8BYpGupBCxAY2cn+RoNEy0Wqvfvp2rkSJJbW3H19MQ9NFRasa1bMXt50bZpEz6TJrH2rbeocXTEf+JEEvbvp0CloqO6Gm9XVwbccguWujoKDx4kp6WFUr2eiEOHCL7oIqLj41EfOSLjv/72n19Ea2urvaLfhubmZqKipK/y82WqAO677z6eeuopQLqif/3rX3niiSeAoxaurq6OpqYmdu/eDcCgQYM4ePAg3d3deHl5odi4caMYNGgQzs7OqNXqX51P0tjYiKurq50Mfe36CSFwcHAgLy/v/M3taGyUPXQBAXTu3cu2rCwSu7oIv+kmsnbuxNvbm69++IHlr72GvqEBJbDbaCQOGcMNAYyAGcgFEoEmX1+uj40lZexYbpg7F9WQIRz87DPWVlRww8iRlABdO3cyJDkZo48P+zZsIPrqq2kuKsLFbCZ8/Hh8bGPjg4KguRlzQwMFViv5+flYs7OxeHkR4+tLgqsrLgMHHi2O7sdxmDt3LitXrqSjo8M+22T48OFERUXxzjvv8OSTT/Lqq6+eVKcYNmwYTk5Ox7X2fPHFF9x2222YzWbi4+MpLS2lvb2dO++8k8OHD6Petm0by5YtY8SIEUybNq2XTGoymTAYDLi6uvLdd9+xZs0ae8AZFxfHddddR1hYWJ++Ebm5uedHqTwWvr7SxfP3xyE8HOuhQzQMHUrE+vW0G41U7N9P8vDhTJ0wgXylkoDcXNqOHMFsMuHp4ECTXk8d0trpgA6gu7GRHS4u7PzuO8pzcli0aBH5RiOlGzeyIyaGQQ4OfAG4d3QQnJCAf2QkptWriRwxgiCVCmV+vlRTa2ulYunnh7qri6T6epIGDKAxIIDSXbsoKivjkK8vYbt2kZSbS0B6OsqzOCfz94r29na7jG9LUn/++ec8+uijwPHLVAHk5OSQlJRk//vWW2/ttd1isXDNNddwzTXXcNttt5GRkcGwYcN4/fXXiYmJAUAdHBzM22+/jaenJwcOHODFF1+0X8jHH39Mbm4u1157Lc888wy33norI0eOpKamhvfff5+2tjaeeeaZPpvcVVZWxrBhw7BYLKxdu5bU1NRzK5wci4QEyM9HERODi8GAWqOBIUPQr1lDRW4uqvZ2figuxr2+nj0aDUFKJfVWK8F6PflqNbFmM+vBniQPBY40NxNVVcXbra3EbdyI3mrFqbkZd42G8NGjGdzQQLO3NwdefRUxcCDDUlNx9fZGOXCgFFO0WqlMVlVJMSU2VnYYFBTgO2IEvqNGkVJTQ/G2bRwxm1lbUUHwG28QkpJC3Pjxv7xmw/8YTrSGxXXXXceLL76IWq3G0dHR7hZ6eXnR0tLC/Pnz2bVrFyArVF555ZVex6tUKtra2nj55ZdRKpXccMMN+Pn5sXXrVjQaDeHh4SiEEGL06NFkZGQwYsQIVq5caV/cYNmyZezdu5c5c+bwr3/9i7feest+8m3btvH555/zz3/+0z5S7+dYvHgxu3btIjw8nA8++KDXC9y0aRPXX389MTExqFQq1q9fD0g/OikpidDQUPbu3Xv+CAfSkhiNsgqluFjGd9nZFDs4sPnBB3l+2zZCnJ0ptVrxMRrJb2hgqBDkId1JAzK2cwDCgWaNhhqzmQAHB6YvWEDznj2Ums3cEhnJge5uTGYzU2+4ASetFo+WFjpCQlDV1ODq48OgKVNwrKyUwoqnp6w+sVql22i1yv67mBhZg2m1Yj18mJrKSgqrq2mprUWl0eA5YgQDBg06YWPk/xqGDRtGY2MjpaWlfXI+24i8adOmsWHDBsxmc686yksuuYQffvhBLuZhNBr56quv8PHxwWq12k/i5+dHeXk5JSUltLW1sWzZMjZs2MCHH37I3/72N/vEoxO1i/zSKjo2zJkzh02bNtnJBnLZqsjISB555JHzSzaQSejGRtm7ZjTKxtHgYKJdXDBOnszs2Fi8fXwY6etLqRDEODmxWaUiRalEDQQCDYA7MjkeajLhJwRKg4G1P/5IWGIibU1NvJGTQ35RES4uLkwUAmtZGVZ3d7Tt7ZSazViamqjOy6Pdz48N//kP+zZulIqqg4O8KSiVsjrl0CE5S6WoCKWLCyEDBpA+aRLpixbhNXYsLfv3s/7991m9ejWFhYW9Puv/NfT1ACFbXm/jxo1MnTrVXhxtI90PP/wAyJ5KUlNT+de//sUf//jHXsuvpqenU15ezrPPPktAQACPP/44qampGI1Guru77QOHTmTdfr5SzgcffMDcuXN77fPll1+ya9curr766l7rxZ33GO5YDBgAHR2yoqNnPomoriY2MZEj7u7UtLWRotfjr9FQrdfjLgT7FQo8AQ8gCFkCFgpsBaIBBeCXn88bDQ0kC0F+ZyehYWHEjhuHeto0An/4AaVCweo33sAlMJDohQsJqKqivKqK3NxcAjMzSXFxwdHHR3aU5+dLZTIpSd4UOjqO9tPp9XgVFjJu4ECMI0dS+sMPVGdmsru0lH1+fkRHRzNgwABcbGsy/I+gu7u7Ty29LaTy9fWlsrKS7777Do1Gg7e3N46OjvbaYzXAI488gpub23F1ZWq1mptvvpmbb76Zrq4ulEolQgi5OLhabR+vcCJL9Eur6IA06bYmvVmzZjFu3DiG9qw+c96qTU4EDw/pvtkmQW/fjjkmBv369VQ5O+PY2soPajWXCMHXCgUBQlDp6opvRwe7genIfJwWmRQ3AlXI9yytq4sub29Ge3uzr6mJoFWr2J6SQnRkJBvXrWMTkFhbS5HJROC0aUQcPsxld95Jt8mEo4MDbSoVuxobSYqMJKy7W9ZVenhI0h05Il1MZ2eZeti/H4euLuKnTCF+5kxqDx6keN8+cg8etK/wMmDAgD5fnPBChV6v79OyLptL6enpyeHDh5k7dy5eXl6oVCr0ej3/+te/uOSSS1B/++236PV66uvraW1t5cYbbyQ8PLzXyaxW6wnvgCcqb1m6dCkrV65k6tSpv7hSzrHDW2bOnElmZqadcBeUhQMZuxUUyN9dXWj0etIjIqicOpU9XV34qFRkFRYS2tCAo4MDnXo95UAI8CMwGNlRkADkAUlAthBgMNDV3o5bfT1ai4V1CgXjd+2ie/Zsig4fJtjdnbCRIzFkZsrReykpRGZmIlJSyMzKwqW+nqyNG6mMi+OWm2+WcVxEhCRebKy0yI6Oknjp6TLVkZ8PHh4EDhpE4KBBpG3dSmlJCYfr61lbWYmHhwcJCQnEx8f3aRfIhQaDwdCnA4Rs57rmmmuoqqqipaWF1tZW9Ho9jY2N9kZXxRVXXCEcHByorKykvLycjz/+mPT09OPya8cGgLZtSqXypEtL2RTPTz75hGeeeYbo6Giuu+46+/Zj15WbO3cut912m30K2O23346XlxfPPvtsn70hvxmlpXIcgsUiy6jc3GDzZvJjY9nx73+zMj+f5u3bOdDYSLrVyvdALDIPF490K8cCBUAnMjHeiuwcL0SSskuhIMTJieS0NIq9vBjn5UW1pyfu7u4kmExM+fOfCQ4KovSbb/jowAESkpMJAaxmM6OvvhqH2lop7mg0knhhYfJaKytl249tynBLi2wRcnWVU8R8fLDm5VFeVcXhri6ampvRaDRERkaSkpJy0sr33zPc3Ny4/PLL+fzzz/vkfLfccguvvvoqb7zxBkIIgoKC8PX1JSAgADc3N+J71phQf/XVV/aDnn322ZMm+o6trbRh165djBw58oT7Dx48mMDAQMaPH094eDgPPPAAIMdAv/322yxfvpx33nkHtVrN2LFje43cu+AsHEgBxTY5uWd5qiyVCr+yMq666iraVq1iQ0sLXtu28ZPVygQgC0m6fKRVWw+MA5qQymUDoEeKKllAuBBs6O6mae9enJyc2BcVRYzFwhEgx2olf+NGLr7zTpISExne0ICptJQurRZaWti4ZAmMH09heTk+CgVzvL1RgFQ0PT2lhdZoZB+e2Sybbzs6pCjk6YkyKYnI2Fgi9+yh2deXwyYTxcXFFBQUEBAQQFJSEuHh4eTm5uLl5fW7dz3NZnOf3kgyMjIwGo384x//sIdDNgNlsVjYv38/KSkpvdeH6+7uthPu5xZu37596HQ62tvbaWtro62tjXfffZcrrriCGTNm2N3BY3Gi4UJvv/02ALfeeutxiUMbtFrt2R0kdKbw9ZX5sNhYjNnZFJhM6FpbGZWWhk9ZGd0aDXEBAVTX11NntaKwWIhCWrd2JLGOAE7ImO5SYB3SAhZoNHRbLPhYreSp1QwSgg2VleR6e+Pe0sLB7m4OuLlhWbUKh+RklhYUMKqmhlZvb5r1eiaNHElMcTE5ublYALWPD1Pb2vBITJS1lVYrebt2UV1by8hrrsHFRkQhZJtPU5Nc5TU+Hi+tlvTSUowBAeQYDBQUFLBp0yacnZ2prKwkKSnpd004IQRms7lPJ3bl5OQA0rPr7OxEp9PR3NxMa2srdXV19oS5uqWlhaKiIrKzszGZTPYNP7dmd955J0qlEldXV1xcXPDw8CA3N5eOjg4eeughvv/++z4bmXdBWjiQLuXmzaDX4+DgwIjBg/Ewm7E6OpKn0eB36BBHnJ2JU6nIMZmIBTIcHbnYYGAbMqbLQZZ65QLdwADACgwwmTjYs82q0VCt1RLa2Eh+VxduFguhCgUV7e2sz8+n5t57ySorw5KYSFhNDUo/P+bcfDNWi4WhOTmsOnCAmuJiNpaVYVm+nLjERMpDQgiKjaVLq0XR1SWtdXw8tLbS2d6O8PFBa7FQmpnJpvJyxo4bR1x4OGnbtzNw4EAqnZzIyc2lqqqK8vJyNm7cyIABA/q0Y/pcQQiB1WrFtw8XV4mMjESj0RAREYHVasXb2xtfX1+CgoLsuWYA9fLly9myZQtms5kpU6bYR2j/nHA7d+487kl8fX35xz/+wezZs+nu7u4zwl1QKuXPMXSo7A7v6iK4rQ2GDUN/4ABdHh7o3NxwbmzEoSeuVSgUKIVgLzI94AoEI13M+J7fHsjG1VRgGNAGpHR2slOvp81kIgAgPJw4f3/EkSMcKiqiXK1mgJsbZquVztBQYlQqvnztNbxdXXHWaIhoaUHV2Mh2jQaVkxOHysuZYDSSPGgQQ4cORQhBR34+Dt99h8rfn42NjVirqpg5cyZmqxXdl19S3dBAQHo6u3U6YpVKIvV6wqOiGD1yJFk5OZSUlFBWVoa3tzcJCQmyEv53IrLYcmTHpsB+6/m6u7tpa2ujsLAQwK7o29zKP/7xjwCoN2zYwJ///OdesdiJCpLLy8tpb29Hr9fT2dmJ2WxGq9ViNpsZOnSofQ5KX+CCtXAgy6sqKuRgn4wMTIcPU9TVRYAQ1Kam4lBaSlVTE9d2d7OstZX4nhjMC9iPdCMzkfWVHkARspXnEJCMFFAMVivjrVbKAV/Aqb4erY8P8c7OBHd0UG8w0GG1EnvoEC5paeQ3NlKxdy9djo4EpaezcOFCohsb0e7Zw7bGRoLi4xkxbx7qlhY2PPQQR9ra8BwwAJW3N6qWFnza2ogYOBCAGH9/0tPT8Y2Px9jZibWmhlIPD0qBEYMG4b5rF6MjIhielkZBWRn5+flkZGSwf/9+oqKiSElJ6VP172zAls7qK+ts6xIAcHd3p6Kiwh4fLl26lL///e/2fdXd3d18+eWX7Ny5Ey8vL4YMGWJflBF6j1moqKjA2dkZpVKJWq3G2dkZg8FAfHx8n/fEXbAWDiAuDpGfT3VEBA2rVmF1cWFYVBSDLr2Ubf/6F42urtSUlzOoqYlmiwUfhYJKIQgGvkd2E3QiibYVWQLmzlF3M6fnMQEUK5VEmUwcPHCAbkdH/K1WlGo1+QYD+7u7mZiVRfTs2VRWV1Ny+DD+mZnceP31mK1WLp8+HVNeHp06HS0xMWTn5LDPyQlXnY5YX1+MQUGsW7eOwYMHEx8TA2vWoIiIIHXOHPtLvXj4cI6sXk1bdjY6lQqXnmIGdXExSUolSZMnU9fZSVZWFvn5+eTl5REUFERSUhIhISF9vpxZX6CxsRHo2wFCAKtXr6ajo4MDBw4QHR2Nn58fzs7OdjJaLBZZS7l9+3a+//57cnNzmTRpEnfeeedxVq6goAClUombmxsajQYPDw+2b9/O+PHjAU5aT3km2Lx5MzfffDPFxcV9cr6+RmtrK+UbNtDi7k6IiwvhEREc/u47Kmtr+b6iApGXR44QtO/cidXZmQi9ntVGoz12C0F2EbggY7gipLuZhbR6WmRyPAqpZHYqFMQKQRWgVyqJ6sn11QLubm7cN2UK2sGD2ZedTWVFBcrcXGoMBiYMHizn1FgsWAwGWv39cfXyYuTIkTgLgbmoiLqmJhJiY7nsuuvYX1ODoaaGsbGxUiDy84OaGkRLC92+vrg0N8systhY+bupSY6l8PMDHx/0Wi3Z2dkUFBTQ3d1tl8MTEhIuqMLp3bt3M3LkSCorKwkJCemz85aUlJCenk5rayupqano9XpycnLw8vKisrJSEq6mpka0tbXh6+uLo6MjRqPxhCUvFouFvLw8u2tpMplYvXo1Li4uTJo0iVmzZvXZbMp9+/Yxbdo06uvr++R8fQWDwUBRURH19fW4u7kRb7WiHT4ccnLIAQ699ho+aWn4+Pry5sqVZG7ahFdgIIaODhTZ2exH1lfakt85yDxce8/jYT2PKZA5unYkIXORzaxJQDHS+gUiJ4VVIadFJatUWA0GvAMDcXR0xNVopM5ioasnH1RTXU0s4OzhwbTbbkNXXEzpvn2EBQcTFBKCU2ws+wsKsFgsPProo4j8fHK++YbQqVPxGzXq6JtgtUJ2NlU7dlDi7s7QWbNkTWJjoySgszPW0FBKS0vJzc2lvr4etVpNeHg4KSkpF0Th9OrVq5kxYwadnZ1npaRtzpw5ZGRkoFAouOqqq1i6dKm9o0b99NNPU1FRQV1dHfPmzWPatGm93hSbpVu2bBkfffQRAE5OTgQEBJCRkUFycjLu7u59Ou/+QovhrFYrFRUVlJWVoVarSUxMJCAgAEVlJeh0VFqtVO3dyz4nJ8bU1aFvaUFTVIQiKIjmrCwyjUaikVYsCKjl6Lj0MiSpsoBGZMrAARiDtHx6ZNyXg4z7bO0+RsAbiAMqjUY6kC5oa3U1QSNHkllWhr6picEWC0ExMfj5+WE0GnGpqcF961bCR49m9a5duIWEYNZqcc/LI9popFmtZus777AjJ4c2Z2dmV1fjl5srJ4P5+cnEupMTFQMHkrd6NW3V1Qy85BLCBwyQCfXOTpQFBUQrFERfeiktbW1kZWVRWlpKcXExvr6+JCYmEh0dfd7cTZuL19cL0YDsgomOjmb06NF2FXTLli32PLM6JiaGm266iQULFhAREcHjjz/OsmXL7ESzWq2oVCpWrVrFnDlzmD9/vv3kd9xxB5dccgkzZ87s04vWarXo9fqTVrGcKwghaGpqoqioCL1eT2hoKBEREUevKSwMNm3CJy0ND6WSiVOm4NfaSuyUKZQfOkTD4cNs0OvxM5vJRybBNyEJVIhMBzghiRWNtF5xSBIWAgOR81FKkbFdBVB9zN9FSBc0AKl4gowFY4qLaW5pwWSxkKHTYd68mShHR3xTUoibOZMmX1+URiPulZVs27gRrZcXLdXVhFuteFosmAYNwiclhYEREYy6/HKZMM/KkvMxU1IgIoKRMTGEhIaSl5eHZ3OzXD9h0CC5VgPIipzCQrxaWxk3ZAgjRowgPz+f/Px8tm3bxr59+85b4XRTUxNKpbJPejjh6HiFTz75hLvvvhudTmfvxLBarXh4eNDa2kp3d7eM4err65k7dy7r1q0jKSmJnJycXoRTKpX85z//ITExkbS0NPsTFRYWolariYyM7JMLt8G2SEhLS0ufJidPB11dXRQWFtLc3IyXlxexsbEnLnatr5dzJHsC8BpHR6zZ2eQplax7+mkOFhZSUVHBESDOwYEqoxEV0jo5Ikn4E+CPtGBtHCWaE7KPLr/n70ikewnSKhYjc3lOSKIeQVo5d62WOKMRhdlMm9WKAWhVKpkSF8ef/vpX2oSguLiYTz7+mOjubvxNJqpcXKhxdiYyMpLBgwbx6Jw5WPR6tuTkoGxqokGlIiw1le6aGqYMHCjXSfD3lxbPz0/+NDVJ19LZWVrE8nJJwLY2uUiJtzdWX19qamrIzs6muroahUJBSEgIycnJ5yyZ/vDDD/Piiy/22dqGtsLl9PR0e8XJyaC2Wq04OTmRm5vL22+/zdixY4GjeTib2b/44os5dOgQ33zzDS0tLTQ3N6PX65k2bRqRkZF9OtvE9sXu6Og454Qzm82UlZVRWVmJo6MjAwYMwNfX9+Svzd8fmpoobWrCsmMH+dHRKGtrER4eWOPiuHHmTP7z5pvENjdzUKslprKSw2YzQUjC1ADDkeJICrCW3nm6eiTpypFWLQ5Zj5mFLIauRgowh5F5PDPyfWtHup3qnmNVCgWdkZHUFxRQlp3Nv9auJcBspsPTk2IXFwJ0OnyFoLOzk3adjgK9nvYDB8j89ltqzWY8hg4lIzcXq9XK4EmT8K2uto+Iz6yqonTHDqZOnYpLfLys0ywslM2xtg4GgLo6lJs2ERIVRciUKXR0dpKdnU1xcTFr1649Z4XTra2tfXp+m6UMDQ3Fy8uLnJwc+9pzKpUKb29ve++durS0lOjoaCZPnsz+/ft57LHHep3MZi6XL1/OU089RUJCAk5OThgMBo4cOYK7uztDhgzpU8LZKlrOZRwnhKC+vp6ioiLMZjMRERGEhYWdmtsRHEz5d9/RFBDAhMZGOkaMoKuwEKegICqPHCHgoosoX7UK/6oqqsxme/LbJpzU9PyUAZcg3UkP5IDZCqTVCgDqkPFfEtISmpFk1QEtSPI5IIlWjXwfHaxWub+LC/n5+aw0GmnKyiKysZFipZLgxETShwzhp59+wtfZmTSDgeZvv+WbtjackpMpCgoiOiqKiyIiUFqtbN6zh+wVK5hw882yHtNkwvmHH/CorUVRWytHPiQlSTfUapXEs1qlqimE7HowGiE/Hy0ytzds2DCKiorIzc1l9+7dHDhw4KwWTvc14YxGI87Ozri4uNDU1MSIESPshQBms5mrr76axx9/HAC1v78/hw4dYv78+aSlpR33AlUqFUIIFi5ceNxC96tWrTpr0v25zMXpdDoKCwtpa2vDz8+PmJgYnJycTv0EHh4MT0ykKzwcn9ZW3IxGthYWomhsZF1REcaSEjKam4k2m1EAo52cWNUzaCgAKOGocLILKa7sQaqWJqTrGU/PmAakiumLTJYXIdMMCiTJQK53kIQktR5J6vTubtyrqthbX0+rRkO7UonWaiW8oIC9jY24u7sz2NeX2qoqKpVKLJmZtGRmYvT2xsHBgQ2dnTjW11OZlwddXURNnEhYYiJoNMRfcgnxBQVyeS0PD7kykEYjSRYSIjstLBa5jkNDg/QKEhKkO757N0o/P+J8fYmbMYPGxkaysrJOWDjdVyLLsQOE+gI26zVgwABGjBgBYB8o29LSYueI0WhE/eyzz1JSUmJvKp0+fTqzZ8/udWdXKBQYjUbq6+vtQ1rVajU//vjjcZXRfYVzoVSaTCZKSkqorq7GxcWFtLS0M16gzzkpCefqaoiJoXzVKsqcnUlsbqYxKgqDEJQB1s5OPBQKtiuVTNdoOKLToUW6k8UcFU48kNOcK5BuYynSLYxAWrZWZA7PZiGrkDGhzQ1t6tlnpNWKHmkBM8xmVCoVkUYjak9PLK6u6Do6ONDdzcCmJpJ9fSlycMBr4ECSnZzIyMjAw2olWanE68gRjDExqBITySwsJLeyEq/ly7nh6qtRmkzg6opITqaxsVG635WVsqOirU0OuR0w4Ogb1dIi6zgrKqTI0jPblIYGyMvDV60mPT0dg8FATk4OhYWFbNq0CVdXV2JjY0lKSjq9m+EJcKIBQr8Fb7/9NikpKUycOBFvb2/Cw8Px9vbG09MTrVZrV/0dHBxkt8DTTz+NVqtlz549PPzww4wcOZLQn800LCsr44033kCr1WKxWGhra8NkMnHLLbcAv7yAwZngbFo4IQTV1dWUlJQghCA2Npbg4ODfdgd1cpKuktWKW2goUVYrXoMHk15fz6fr1hHs789/Dh8mtLOTiIEDEU1NNOzaRSdHZX8D4MnRVEEOMmc3ECmIlCFJKZDEK+aowNKMrF6ZhHQ9TUiBxRkpqhgBLBZygMD2dkJ7VDR3jYY8FxeKFAoG1NXhqlIxb8kS6urqcGtsZMiwYRTrdASYTDg2NxMTE4PRaCQgOprK8nLCExLAaCR3yxYqjUaampoIdXZmXGysVHHb26V4Ymtq7u6Wlm/iRGn5WlrkeAg/P0lQqxXy8nAEBiUnk5qaSmVlJdnZ2Rw6dIgjR44QGhpKSkrKGa3t1traSlVV1W8m7bFYuXIlarWab775hg0bNuDk5GQXG/V6Pffee699up36rrvusqtDtoa8E93lXV1diY+PJyIiAqVSiY+PDwMHDuzTCz8WZ8vCtba2UlhYSEdHB4GBgURHR/fd3S4+HvLzCRg8mABHR0hOpm7TJo40NWHOyCDJ2xtFbCwebW1sbmggCFm43I20WM5IAhUjrdJAZN3lYSCt52+bJVQg3UU9cHnP40akKxqMJKBtqIUjsk6zEklKd6sVK5K4nQYDyUOG0NzSwuH6eppcXelauBC3+noOeXvjaTaj9fNjxebNXOrmxhitlnBXV3wDAwmZNEmO6wNCnZzo3LIF19JS2sPCjlq14GBJskOHpFUbOfLouIqEBBnXZWbKbUOHHl1WWQgoLUVZV0d4XBzhl15Ke3s7WVlZxxVOx8XFnfLN8pNPPqG4uBilUslDDz3EM888g1qtprq6mvj4eDo7O3nppZe49957ex3X0dFBcHAwXV1deHl59Wofu/vuu7nkkku46KKLeOutt9BqtTg4ONDe3k5ubq49B6dSqVDcddddYvLkyWg0Gr7//nv27NnD5MmT6ezs5P777z9u0GtGRgaVlZWYTCb8/Pzss9j7GpdddhmzZs06Lm48UxxbJeLm5kZcXFyfr08OyOWnNBpy29robGmh08WFgu+/5/P9+xlpMHDTAw+wePFiMqqq6DKZCDaZKOBo/s2ItG75yFrLUGSMp0amEPKQRAlFuo5ZPf8nIgWV1p7LCETGgkU9/7sBg5DWbqNSibnHwmmANFdXWi0WjCoVQSoVZUYjJhcXRnp709XZSZWLC65ubowOCEDr4kLosGHcPmMGGq0WoqLk0rrl5aiNRqlMVlYeLQFTKGQPockkxw0WFcl4LiREFoLn5UklMyREupXNzbIzPTBQ7hsaKo9tbpZxYWQkZouFgoIC8vPzaWlpwdHR0Z7TO5XC6eDgYDo7O1GpVPZKGL1eT0lJCTNmzGDRokXHEe6KK65g7969VFRU4OnpyZ/+9CdeeOEF+/Z77rmHhoYGPvvsM7777jvWrVtHaGgoixYt6nVDV69YsYJNmzbh7u6OWq0mJSUFnU6H0WjsFZdZLBY++ugj3n//fVxcXNi4cSOPPPII1dXV3HjjjX0+8ryvLJzVaqWyspKysjKUSiUJCQkEBgaevRF8gwbBgQO0ZGWha2hg/H33kejqivuAAbQ1N/PB8uV0ApMVCjLc3TG1teFhNsuFQZAEKkUSKhdJInckcUqQxMzq+T8Mmb+r69k3HEmyqp7jHJGqZz1SBd2KHO2QYLXSAtQpFFgUCmqR8WGAVkue1UqLwYCXgwNljo54+PoyoaWFRE9PMhQKCnQ6Ap2c2NfZyai4ONi3j32rV1MXE8OMnhYUEhIkqfbvlyv7jBolazNBjnoAmTooK+udLLfl8/LzZd9hZKTM6Xl4yOP1eli3DnV4OElRUSQlJVFXV0dWVhZ5eXnk5uYSFBREcnLyL4YIVquVwYMHs2nTJvtjTk5O9l7QE2Hnzp384Q9/AOQ0u1WrVtkJN3r0aLKzszGbzSQlJVFZWWkPh9599102btxon/Wqrq6uPvEz/AwNDQ28+eabbN26FScnJy699FKefPJJRowYYV/cri/RFzFcU1MThYWF6PV6QkJCiIyMPDeVKwEBjLjySszV1bBlC3vMZrx1OooMBkqysyno6iJXqWRaWBjbW1qIBjYgCRXV87sb8EMKKo5Ii1WLJJpNLKngaA1mBVJY8QIm9/wtkB3lIUh3FWR1S07PfsOFwKpQ0OXgQLWnJ9WAqqWFWLMZXVcXQqMh0c2N9Ntuo6S1lYBt27g2PZ1NeXmoVSoGOTvj5OeHw+WXE1FcLBe1jI6mprGR2q1bSRs1CuXQobIwoLFRqpOennI/f385S7O1VVo5tVpaNdvyy/Hx0q0sKZEjIUD+vugi6caWloLBQICXFwEXXURXV5ddZFm3bt0vFk4bDIbT9m66u7vtU+j8/f3Zt2+ffdvu3bu58soruf/++xk9ejTjx49ny5Yt1NbWEhwczIoVK1i0aBEA6ubmZnsHgMlkoq6ujv3795ORkcF1111HamoqIO8ARqMRJycn6uvraW5upry83F7C0tcW47dYuK6uLoqKimhqasLLy4uUlJSzstLlSREcjConB9XAgXKyb3Y2pS0thPr7sy86mtFmMz90dJBfW0uTuztVra1cgiSTHzLOqkXGat1ICxaKlPubkPGazRrW9vw9FUlUkGP5YpAuqgVJPmeOxnEaZFqhDHDTaHDp6KCitRWzEGi1WnwDA5nl7097czM6d3feXrECo9FIXFwcawoL0W/ciD4nhy+cnBg2YQLRcXG4Dx0qCbJ3L80bN1IbEkKkuzteIEURkKP7MjJk/i4gQD5mG/VQWCiH2IaGStcRpDsaFSUJarVKolVUSMsXGSmrVw4ehOZmXDQahg4ZwuDBg+2F0/v27SMzM/O4wmmTyWQvtLct5JGenm5fY+BEcHFxwWacamtrexXqK5VKPv30U6lCqtUsX74ckCvsqFSqXhU06nfffRd/f3+EEPaZJd3d3fj7+/eq8nB1dcXT05POzk68vb1pbW3ln//850nnkvxWnMlcE4vFQllZGRUVFTg4OPx6lcjZRGAg1NYSMX06Lu7uuEdGUvT557heeSV7168nSa+nsLOTsQEBbAQOt7djtFopAiYiZX4TcuhQCTLJHYMUU9qQRJqBFF2Ken5s7qa153cE0pJVIImrQ4ovRiDHzY12nY46gwF6jrUCbhYLDl1drKipITgkhNjubpxKSihqaUFXU0OalxdFjo7sqqjgT/v2YSgoIGHyZMZPmSIHFUVEEH/ffQRkZ+NVVyfjL9sQo6AgWYvZ0SGtmlIpiWcjkW14bVmZdB+FkERLSLCLM+j18tiqKumKjhkjHzeboaAApdVKdFAQ0ZddRktLS6/CaT8/PxISEnrNM/klkh2LkSNH8tVXX/HKK6+wdevWXtqC2WwmJCSE6OhozGYzV111FWlpaSQnJ2M2m+3uJIDi3//+t9i/fz86nQ5vb28GDhzIxRdfTIDtDnQMdu/eTUBAABEREXz66acEBQXZV4/sSyxbtozvvvsOk8l0SmPMbFUixcXFmEwmwsLCCA8P7/NUxWkjJ0dWXVRU8PayZZRUV/OHsDDChg3jy88+Y1luLlqzGTelklX79xPW2Ukhklw2uR+OKpQgrV4zUhxR9uyX1XMMSCtWhSQlSKs4EmnlipDEUyJzdu0cTZZHIvvw/AMCaDaZONjcjEKhYMCAAVjNZkJLS1ErFPhOnUphYyNGo5G5c+cyatQoktvb8bBaYcQIufbBscjMlHFccvLR1IAN+flSYAkOlm6kzQuxWqVVUyjkj6OjJKRCIRXP4mK5f1OTTMX4+MgYr7FRjr/QaiUx3dwgOBij0UheXh6ZmZl0dnZy++23s2jRIvsCpDb4+/vbh2iNHTuWLVu22Bfy6OjoICgoiO7ubjw9Pe1NrAAxMTE0NTXZdQ+TyWQfryCE6NV3pxAnyVifSAQpKCggNDQUZ2dnjhw5wtatWxk1ahSDBw8+0SnOGP/617/485//jBCCn3766aSj+EDKtQUFBdh6+mJiYs7PQo4ngl4vv2xRUby/eDE7Wlt59b772FVWxs4vvqBxzx5+6hFNio1GWrq6SDCZyLRa0QhBFFKVVCKJdKTntMeSjp5tZUhXFKRlC0O6ogYk2WKQlq3imMvz4OgclUM925VKpZy9YrXSpVIR4O3NiIEDyezupk2nIwgYn5qKJiSEBx5+GGV1NYqICMzOzqiamlC0tEiRw81NWqqwMDkNrKtLWjKlUpKjquroZGiQ2zo7pdgC8kZlEz2MRhnLVVVJz6FnxL4dDQ2wd69cm33kyKMrwBYVSfLGxdnzrZWVlSiVSt577z3mzZt3hh/smeOkyYufD4EFuPrqq8nKysJqtXLHHXewbt06lixZQmtr60mfQKfTMXLkSLRaLUeOHDluu9ls5qabbmL8+PH29QXmz5/P0qVLCQoKOqGlBemH5+fns3fvXkwmE6mpqaSkpFw4ZAOZDO/qgsZGZt12G3+7+25EZCRdWVkYIyPRx8QQ6uyMl7s7k6xWnMxmSoBIITAiCROGdPUKkGIJyBjOE2m9QLqVMcAIZOpAjSSnAzJmA2ndWnrOoUVatEBkvFfVc5yzUonVakVntdKmUjHAy4uxCQk4+vvz4MMPExoaSpXVyqd79uBRXU32p5+yKiODNquV5cuX8/7q1XRHRMibzK5d0mLZxAkXFymEWK1SobTVWtoQEiL/V6uPiiI2W2CxSNd02DB5ntxcac1AJs7r6+X66EOGSKuZmyvzfm5uEBdHR0cHCxYsoKqqijFjxhAUFHTeFjI5JcnOZu1scdH27duJiYnhgw8+YMSIETQ0NJy0qt/Z2ZnVq1fbB8H+HKtWrSI0NJQPP/yQBQsWsGPHDsaMGYOXlxfu7u7Htf4IIaipqaG4uBghBDExMRfs7AxA3o23bOFITg7Ojo5U5OfjHBzMTdHR+C5ezPa33+aRTz9lU2EhA5VKmoTACylqtCCJ5Yy0TE3I6c21yMR3EtKVbETK/mqkolna89StyNKwZnpGNfTsP6zn/6ye/WqQxc3xKhVqnQ6DQkGXiwvl/v50KhTUbdhA/p49BMfH09jSwqgBAxi6aBG4uqLfto3DX3yBtakJDy8vSr/7jqTLL5duoF4vv/xKpbQ+9fUyLZCQIJ+4okL+2L78iYm9rVp+viSvv7+M/eAogRsb4YcfZHx4rFXz9ZWLVg4YACoVmzdv5tZbb0WhUPDTTz8xfvx4rFbrhU04G8LCwvj8889ZuXIlc3oGzQQEBPzixC61Wv2LJTgZGRlcfvnlgFxlx0a4E6mUbW1tFBQUnJ0qkbMFhQJiYxEtLai7ukitrcUyeTIRQEVDA/lCkObiQk1AAEYhUCuVZLa0ENLdTRVQrlCQLAQRSOIUIEmYj+wqCOToFGeQ1msA0uoJpEsaBQxFWrnCnnOokMnyNqBOqUTo9RitVpoBD7Wa1IgINvYsVWaxWOg2GIivr8fZZKJDoWDl6tV0d3cTERFBZnk5iydNoqOxkaC0tKMj1Z2cJLmysyV5fH2luGH7zIKDpQCiUklrVlEhR7SDJKHZDIMHy6R3bq59dgptbZJUF10kj8vPl/tbLHL7gAEYDAYef/hhXn/9da6//nqWLl1qT4orlcrzdoM+JcLZLu6FF17g8ccfR6lUcumllwLwxBNP/KbGwdbWVntO5NhVdo7NwxkMBoqLi2V9n5sbQ4YMOTtVImcJRj8/XNra8JowgSiQwb5WS/GaNZTX1RGemMhNAwaQ39iIf0MDxspKKqqrietZTfWgyUQc0kLZfmwDiWqRBLQVLncgLVcy0tp1IsWRtUhrZ0uUW3qO9wamC0GNUskBlQqLxUKHVktbYyOB7e2YAFeNBkdnZ3a0tWGxWukqLUXf0IDK3Z2uggKCtFqMqamEBQbKTgGbVXNzky5fYuJRlbGsTAofVqu8GSUlHbVONgWyuloqmrbyMFsHS0MDrF0rrdzo0Uff4IAAKb7Ex4ODAwcPHuSmm26isbGRL774gunTpx83Rfzee+9FqVQSEBDAp59+elbGLZwIp0Vzd3d3PvvsM7766isMBgNvvvkmTk5OJ3Qnly5dysSJE1m6dOkvntPLy+uEq+zYLFx5eTm7d++mubmZhISE3x3ZQCZNqxUKWo8cod3Hh6379lGgVJIaG8t0Hx88fXwwVVSQl5vLhtpaglpa0Gg01Pr4kIKMpwuQ8Zwz0K1UUoYkFUjXsQxZuhWPlPiNSKtnRpJSIAlmQrqiLj37egNrkFUnyYCbWo3FYqGitha1qysh7u44Ojri2DMW0dvbGyd/f8xBQfhZrYQplTi4ueFvi7UdHaW8bzBIcqlUMv6ywRarqVSSaJWV9ILNqmk0knxtPXpre7tUINPTZXVKXp78yc2VaYaUFMxKJc8++yxjx44lPj6ezMxMLr/88uPEv5CQENauXcvmzZuJjY3lm2+++Q2f7unhlCycrQn1L3/5CwkJCdxzzz28+uqrHDx4kM2bN3PPPfcwevToXsrmn//8Z/785z//6rlHjRrFjz/+yIQJE1i7dq29+0CtVtPe3s6aNWuYPHkyUVFR5+wu1Nfw8PBg2pw5OBQV0WwwsLWsjJS1azGHhODr6Ym7RkOtoyN3REVxsLOTfGCKszO1Li6U5+WR1NZGtsVCLjDM2Zlsg4EupKQ/Hek2grRkTvROCzjRW8XsQCbXR/Y8VgwgBJUWCxU9ymhEdzf1QLleT0doKLW1tWibm4lSKHB2dia0uxuTwUBrSgoO/v6MGz4cRW6uFDxsQlFKylHLVVoqrZeNaMdatc5OSZr6+t5WzXYTr62Vs1Q8PGD48KNvalCQ7EKIjgYXFwoLC7n55pvtkwvmzZt30vzrsXkxW1vaucJpWTiDwUBERARCCEpLS3n99dcJDg5m//79wMl74i677DJ+/PFHFixYwIcffghgTxzOmDGDiooKxo8fj7OzM6N7XAUXFxcSExO55557uO2229i4cWOf99ydSzg6OqJITMS7oYHkESMI9PYmLjycoEmTSPPzw3fMGLLd3UmfNo3g8HA6iovxyMkhwGzGXaViikpFsoMDCicnJlitJCOt3XccFVWKkbGbO9IaguwmyEbGcRchOwnygI1IBTMJ2QQrhMAd6YYeMZkwA4EWC64uLri7u9NuMtEiBMEuLnhrtWiUSjp0OhYsWMCEqVNljq27WxJEiKPlWCDJYbEcFURqa49uU6kkEVNS5HE9Da6AtFxNTTK5nZgotxUUyKqUlhZISUE4O/Pmm28ydOhQXFxc2L9/P3/84x9PqdihvLycdevW2TWEc4GT5uGOhc3C/d///Z995RGLxcLzzz/PQw89RHBwMPfcc0+fDoO1IScnh2effZYVK1aQlpbGgw8+yOzZsy9cVfJX0HLkCIf37SNmyhS8q6vZq9ejbmtj97ZtFHZ24tnayroDB8ivqSHUZMLi6UmDwYCPELhrNGTrdOh0OhI42m4DMj6rA7odHWWtIJJ03Ug3pq3nJ5reFhAkCQchxZQCjo6PE2YzA52cUFksqJ2cqBGCxDFj5ASuw4eJAkaMHs34CROobWjAFBoqc7JCyOS02Sz/VquPVpGArJ+sq5MxWUDA0YJmGyoqpBDi6Slbdmzo7JS5tchIcHenurqaBQsWsHXrVp5++mkWLVp0yt+L9vZ2ZsyYwbvvvmtfu+1c4JSuzvYi5s2bR3NzM5s2bSI9PR2QFmrSpEm99utLJCUl8fHHH5OTk8OgQYO44YYbGDx4MP/+97/7dD2Dc4UODw9ETQ3Gw4epraqCykrqzGaC3d0ZkZqKY0ICI/z9SU1JIWLiRGKdnHAAas1mqhsbcetRbvOQRLENxMhDWrZAs5lApHppRHYPNAP1CoXd2rkiRRfHnt9qpKXsQlo8LxcXHB0dUarVGE0m2oxGvH19GZ2YSGlPL9oNCxag8PenoaKCLWvXsn/nTirLyuTFKBTSqpnNR9XGY9cddHSUj9m6CnJzj1rEzk4Zr40aJYmVlydjwbIy6XampiLc3Fi2bBmDBg2iubmZPXv2cPfdd5/y989isTBv3jyWLFlyTskGp2jhToazYdF+DbW1tfz973/n/fffx9/fn3nz5uHo6MhDDz10Tq/jt6C5spLsjRvp8PQkOD+f6HHj+PrHHyk/cgRzcjKHMzOZ4OvLqFtvZevWrRR//TXrGxooLCrCz2JBgcydOTo6EmI0ohcCZ2RuzhMpkOw75vnCHR3xtFo51CNe2EbuhSCJeqxsoVarGeXnh79eT2F7O5VaLVaFgpSUFCaPH0/9zp0kDxhASmIi23Nz8U9NRavVUlJUxPz0dAJsKSBHx6NFy3C01621VaYHYmKObhNCWq6yMin99xTMAzL2y86W5/Lyorm5mUWLFvHNN9/w8MMP8+ijj552DPb555+zaNEiBvYsYHL77bfb01xnG6dNuGN3Py9FwT2orq7m2muvZevWrWi1Wh5//HEWLVp0brsCzhAVFRXkfPcd+W1txKWlcXFICKvy8ujct4+dK1dSZDIxLDmZGWPHUq3R0KrXYzp8mG+PHKEoNxd3gwEV0OroiEKhwLFnIJGtbEuNtFwlgEGtljWlFgupZjMqJFmreiZKuQFRGg21QtAM+CuVDA4NpRpQNzaiUKnoCAxk4qRJcp5IQwP1BQXEJCRQXVeH94gRjBk7lqysLIJdXRkRFCRjNZXqaJMpyLispES6iR0d0s2MjpbWsKtLuqCRkVKV7OiQ+xmNUu2MjUUIwdq1a1m4cCFubm588MEHDD9WRPmd4LTlmfNJsmPh4+NDcHAwmzdvZseOHbz++uu8/PLLLFy4kHvvvfe8DZA9FYSFhRF8662M278fVWoqZfn5uLa1scVoZHdnJ/5BQVT6+LCyuJjcQ4cITEjgokmTSGlpId9gICM3F3/AV6MhbOxYuTBmWRnJKhXZFgtCpaLU2ZlQoxGLQoFCrcZNpbLPOgl2cEDv5ERdczM6oECtJhaYrFaT5+hIS2AgA2NjWbVqFZjNjDObCamvJz8vjy4fH6ra2+k2mxk+dSqasjLCOjvx9ffHPSTkaOIapORfVSUJ5enZe5iQySTjNFvVv62SRKuVRNu3TwolYWF0dXWxePFi3n//fW6//XaeffbZC6uE7zTwm1zKCwl6vZ733nuPl19+mZaWFm655RYWL1580lrMCwI6Hab6etYcPsyRNWvILC+ntquLu6dNI3zMGP6zciW7V6yg3Wpl9i23sHPjRqxFRexobcXBwYE4b28mpKWxrriYvLw8nIAYq5U8wEGlIshsxkupRFgs5Dg7Y1WpMBgMqJRKIk0mjECLoyMBXV3o1WqUERGEmkyE+vlR5uBARXU1VpOJ60eOJDwxEVN7O21dXTgkJuLf0zUyOiYGTUuLtGpKpXQVbekbnU5K966u0lK5u8vYDqSiWVQk1c2GBhnDBQfLYzo6ID6ekpIS7rzzTrKzs1GpVLz77rtnpTvlXOK/hnA2mEwmPvnkE1588UXKy8u5/vrreeihhwj/eWvIhYKSEnJbWli7dSuO+fn4pKeTPnEimR9/zAvff0+7TkeYwcCQyy7D5OjIfz74AG1jI3GzZuHt7c13y5aR6ObG3s5OtEolXp2dRPdUjhS5uqJWq6msqCDF0RGDRkOl1YpCocBHqyVWr8fbamW31Yrw9sbb25v29naUCgVxQuBrNnPJ9dezsaaG0LAwoqOj8XR1ZaCrK3v27KG2ooI7n3oKtY1ENnXSZJIEcnPrbfHa22VtZEODJFd09NFtZjNs3SqT2l5eGI1GFi9ezBtvvIFCoWDEiBFs2LABR1vZ2O8Uv0tt/Zc6EDQaDTfffDOZmZm899577Nmzh8TERG666Sby8/NPcsbziKgoEltbGevnh8rRkbHOzvj7+3PIYsGnro7Jkycz4777GBMWRmRgIAOGD6fN2xtNfj5Hdu4kWK2mvaWFKV1deJvNOKSk0DZmDKMuv5x7Lr8clUqFSq2mUKlEr1AwXKkk3cuLVB8f8rRa1qvVBPj4EG82093cjMViQdPejp+PD44jR9LV3Y1TTQ1OTk5ceeWVGIWgorERjVKJf1AQZZmZ9qp+AdRbLAizWVq57m4pktigVktiJSZKUublSZWytlZau4kTwcuLI0eOMGbMGL744guWL19OXV0d99577++ebPA7tXC2ibYP9EzAOnbF1p/DarXyww8/8Pzzz7Nz505mzZrFY489hqOjIwkJCRdGTNrVRWtODj8WFtK5dy+TUlMJHTaMF/7yF3SlpQSPGkVgYCDe9fV8X1BAWVUVps5OPLq6yHBywtXNjYb6egY7OKAdMICA2FjCwsKw6HSsf/ddSlUqBsTGYq6qor69HWE04uTqSnFPg6RtjeoBCgXjw8Io8vDAITwcrVaLRqMhJjSUNDc3BqekUFFTQ41Gg9LdnebmZpw0GmYmJ6MA8nJzqdfr8UlNJdnWs1ZbK0nX2SnjuGPVSbMZNmyQuTYfHywWCy+99BJ/+ctfmDp1Km+//Xafr1J6vvH7WAX9Z/i1DoRjoVQqueyyy5g2bRpbtmzhr3/9K0OGDAHgo48++sUSoHMGFxc8AwIYbDaz22xmW2kp0T4+7DYYaNbpuKirixvuuIPmpiaSX36ZBg8PXD08KCsqYqbFgs/Eiaz67js6nZyIBMwVFRj8/XH28MA7IoLBBgPXzJ1Llb8/jz/+OLW1tUS4ueFXWEitQoFrQABpDg44envTMXo0t6ekoPXyYlV2NjqdDt+gIHTV1WzYuZMgPz9cNBoCU1NJS0uTSzC5uEBtLSHBwegKCgg/tpjdzU220vj6SotXWSnnltTVyTTB1KmgUFBSUsL8+fM5ePAgr776KjfffPP5/1zOAn6XFs6Gm2666Vct3M9x+eWXU1dXh6enJ5s3b2b8+PE8/PDDTJky5fx/wFlZiMREvl21Cp+6Oo4YDLR2djIpPh6Niwv7KysxGY2os7PRjhjB4dxcBqWl0bptG19lZuIdGsrQoUOZEhGBa10d+UYjqw4doqmzE0N9PdMGDuSd9eupaWnBycmJ2upq0h0cmDR0KBsVCqoaG9FqtcyYMYM/33knLfv20VFdjcnZmW0NDfgFBJCcnEyYry+ujY3SbTQYJJlslujYOM5qlR3dx+bj2tpkbeSgQeDrixCC999/nwceeIBBgwbxr3/9i6hj9++BTqdjypQpZGVlsXPnztP6zC8k/K4s3NKlS1m5ciUzZ848pcLoE+HDDz/Ex8cHhUJBVlYWzz33HDNnzmTQoEE89NBDzJw58/yVjSUno8jKYtasWSAE4w8fhrQ0MjIyKNmxg7qqKnQaDW6BgSj27WP8pElcOmsWWwMDuaSqiuioKAalphIycCBrtm5FU1zMgIQE3v70Uzo6OvCKjCTQYiHYy4vRw4axceNGyl1dWWkyoS4tJdjREVPP0tMr/v1vQpRKLrrySmhrw7mxEVV4OBE2EUSvl9UjDg5SVfTzOzqDJCBAqpMKhSRkW5ssPrZZtcmTQaGgrq6OhQsXsm7dOp588knuvffek86h+bVG5lNBTU3NOVuD7mT4n7NwJ0JJSQnPPfccn332GbGxsSxevJhrr732nK++WlNTQ+a2baR6ehI8caK0FgUFdEVFcejQISgqwuzmhsZioaGsjMFaLSHDh7Pz0CEyGxtx6Oxk986duCYmkpiYyJ49e4iwWikqLaXNw4PK0lLc6uvxd3MjPjmZz7ZsoRUIDw/Hz8+PoampOBYXY6mtJXTqVNwTE/H09GTs2LFs+PprRHU1Kicnpk6YgCYy8mifmtEorZrt/XJw6D0wqKpKFhwnJ4OfH0IIvv76a+68805CQkL48MMPT/kzPJPP3GQysWjRIqqrq7njjjuYNm0acPzcnr4eZnwi/C5VSjhxB8KvYfHixYwfP5558+b1WqUyKiqKd955h/z8fCZOnMiiRYtISUnh7bff/sXVLPsaVquVvOpqDuTny7kf5eVgNOKSkcEoX19GjRzJOFdX/AICUMbHkxMZSWldHXUWC77h4Uy79Va0QUEkmM3k5uRQWlqKITgYq0rFwLIyBru44DJoEJW+vmxsaOCSceOYk5pKTGQkf3/hBeaNHUu3QoHThAlcNnYs2s5OtmzZwrvvvsu63btZ+9NPVBcUoFCpZDxmg4ODHO7T2SmT3LZWHJBWTaeDCRPAz4+2tjZuuukm5s2bxy233MKOHTvOqnt44MABxo8fj7u7Oy+//HKvXkpbH+aBAwewWCznJKT4XVu408GBAwd48cUX+eSTT3jmmWeIiopi7ty5J9y3qamJpUuX8u677+Li4sJdd93Fbbfddk7Kxn766Sfa29u5auBAKS64uEjRoavrqNUoLKTd0ZGfdu+mqamJYb6++IWFYfT2Zvny5WidndEUFFDd0EBkVBSb9+2jRqkkSaPB09+fPXV16PV6xowZw+pVqxiq03HHbbfxXkEBRquVhx9+mLi4OFZ9+CHthYWER0RgNZvpCgkhISWF6Oho2R5TVyeT2TqdrBCxjSzvWQGHlhbZCdAjcK1fv54FCxbg6OjI+++/b2/FOh2croX7+9//TnBw8HETuoQQLFiwAB8fH1paWoiKiuKRRx457es5XfxuLdzpIiMjg4svvhg4OjvlZPDx8eGZZ56hsLCQP/3pT7z00kvExsZy3XXXcd99953V65wyZQqzZ8+W4wJsq4f6+tLQ1kbO+vVyp9hY3I1G1O3tuLi4ED5hAmHh4ZRt2oSorydJo2HomDHMnjmTiKQkosaNY/bs2WiSklB4efHElVdyx403cvWoUcQLwQ6Lhc+Ki0lQKBgYEEBcT7tMh9WKWq1m3JgxpE+axLT0dEk2kEOBfHykFf75Wtl1dfL36NHg50d3dzf33nsv06dPZ9q0aezbt++MyHa6MJlM7Nixg7CwMHsNcE5ODjfddBP79u3jxhtvpK2tjXfeeYeqqip279591q/pdyWa/Ba0trbaZ8MfOzvll+Du7s6jjz7KH//4Ry6//HKWL1+Os7MzGo2GBx544IzWJ/s1KBQKVCoVhYWF5JeXM7SujoCpUznS2IixrAz3/fspaG9n2LBhTBw6lLUbNrD/iy+YPGkSyvZ2hrq745CSQml9PR4eHvhqNPjt3o2uZ0kmR0dHDldUMECtZv1PP3HZffeRWFKCn58f48ePp+bwYfZ89BGhvr6MjIigbfBgFLYayOpq+RMYKCtGfH1lGw3IKpLDh6VIEhUlu7qBPXv2cPPNN6PT6fjmm2+45JJLzth1u+yyyzh48CB5eXksXLiQm2666YT72WKxrq4u+wIfCoWCw4cPc+uttxIREcH69euZMWOGfSmp4cOHn5P629+Nhfutnu/JZqecCry9vZk8eTLl5eW88sorrFmzhtjYWBYtWkR5eflvuq4Toampic2bN3PoyBF0QUGQmckwX19SYmLoPHgQ9Z49tO3di4ezM6leXrR2drK7pYXxd9+N25QpkJVFfHAwgYGB7C8qQpmSQvvBg3jk5eFdX49nWBgHvLz4IieHXR9+SCBQWFhIYWEhrmo1GzZu5Ifdu4kKDGTwsSvKBAfLvNqhQ1KBPLaAWKeTsVxPEttsNvPkk08yYcIEUlNTOXjwIJdeeulvipO+++47qqurycjIOI5slZWV3HHHHXR2dtqfw8PDg0mTJvHxxx8DMHDgQDZs2MCDDz4oa0+dnCgqKuKxxx4jJyen1+iFs4XfDeFsb+KcOXPkAhmnCdvsFIC1a9cyduzYUz5Wq9Xy0ksvERISwvz588nMzOSdd95h586dJCYmcvPNN1NQUHDa13QyeHh4kJCQwPXXX09sSgqEh+Pm4EBIejptAwdS5+BAO0BcHIk33khIXByte/fS3NzMyDFjSLrmGtKCg1n34YdUlpUxxs+PqKgo/AYNQmGxoDObaWtro9tsRpmSgt5iIbyqisEWCxYXF9xGjMAxIQFjbKy0aLm5R5cKdneXg3ySkqTcn5kpLZurq2woVSjIzc1l/Pjx/POf/+TDDz/k008/PeOlnE8V27Zt46233uLRRx/lrbfeAuRNetiwYdTU1LBnzx5ArpFhMBjo6OggLCyMESNGMGLECP72t7+dk+FUvwvRxOYibNiwgTlz5tgX+TjdBtjFixeza9cuwsPD+eCDD37zTEur1cr333/P888/z65du5g9ezaPPfYYAwcO7HvFq6kJOjr4LiuLXbt2ESwEfv7+TJw7F61Wy6716xHFxVR7e6Pr6OC6KVNY/+676EpLKYyOJq+oiOHDh7N50yZG+fiAkxO7amp444UXcGlqok2pJCIoiC3ff0+Lnx+XXHaZXN3WNmektlZW+sfFHZ1NUloqayF7yrXeeust1q5dy7p165gwYQLvvvuu3Y0/29i7dy9z587lzTff5JFHHuHGG2/k1ltvpaWlhb/97W9UVFSwePFiNmzYwMqVK3niiSeOm2VyLtICvyvCjR8/nmuvvZY777yz1zY4e316p1LhIIRg8+bNPPfcc2zcuJFLLrmExx57jJEjR/bJdel0OhYuXMiY2FjumD8fnacnn3/+Oc05OUwcNYohV1xBR0cHW77+GpfMTKpVKjxSUjhQVoavtzfedXVkNTYy/6GH+Oijj/D39yfZzw/zjh0kXHop2hEj2L59O+Hh4YSGhEBBAQHBwXICV3u7JJlaLQWcggKZd1OpZCdAj3JbUVHBzJkzOXz4MA4ODrzwwgv2NdHOFa688kpuvPFGoqOj+fvf/46LiwuLFy8mNDSU1157jezsbLq7u/nnP/+JT8+iI+eCZMfighdNbG/Irl27qKysZNq0aaSlpTF37lxmzJhhL5I9W2/cqVQ4KBQKJk6cSHp6Onv27OG5555jwoQJjB8/nkcffZRJkyah0+nwsCWKTxPd3d1YrVZqrVbySkpIjInh0ksv5QeDgbJ9+3CvrSXl8svx0GqpHT2aBIuF4Lg4HIOCUCqVTLztNm5ob4fWVp5+7DGMBQU4eniw2t2dL3/8kZFVVXjGxuLv709AYKAkUXGxbKMJCTma0La9v56ecpEO5Pv+ySefcN9995GUlEROTg51dXVnLX9pNBopLi7uVXhuu+mOHz+eyspKZs2axSOPPMK2bdu48soreemll7j//vtlJ0RPr55tMNa5Lue74GM425u5ZMkSFi1axIoVK7BarYSHh3PVVVfx5ptvAthnxX/xxRe8+OKLdoHkt+J0CqVtfVtfffUV+/fvJyAggMsuuwxvb2+mT59+xvPs/f39ee+995g6dSrVFgv1jY1ElpczPCyMuuBgvKdPB4OBlEmTqKuro9HDA11XFyM8PJg8eTI7d+6kuL4ezGYU27fjGBICkZEMGDAAh5gYQi66iEkhIYSZzUfjtbQ0qUB6ecllt7Ky5O+YGDvZmpqa+MMf/sBtt93G/fffz8aNG4mLi2PcuHFnrVH0lVde4bXXXqO4uNj+mEKhQKFQ4OXlxRdffAHI4VM33HADWVlZLF++HIPBYCeb1Wo9f0uZiQsYVqtVCCHEkSNHRGhoqCgtLRXjxo0TBw8eFEII8cQTT4gnnnhCCCFEUVGReO+990RkZKR48sknxaOPPio6Ozv77FpuvPFGcfjw4dM6Zv/+/cLV1VUkJiYKFxcXMWjQIPHJJ58Ik8l0RtfQ0dEhjhw5It+X+nqhz80Vubm58n+TSRz897/Fu6+9Jh555BFx6aWXimf+7//EwVdeEauefFIcWL366Ilqa4XIyhLCbLZfZ9G6dULs2ycfr68/uq9ef9xjVqtVrFy5UoSEhIi0tDSxf//+M3o9pwPbe2YymcT8+fPFBx98INrb24UQQlgsFvt+kyZNEnv27BEfffSRiI2NFU8++eRZv7bTwQVv4QDuuOMObr75ZrZt20ZERARpaWk0NjbS0tJCUlISJpMJk8nEo48+irWnUuKJJ57AxcUFkO7D+UBqairr1q0jJyeH/Px8JkyYwB133MHAgQN55513TtvtcnV1ZcCAASgUCjqcndmZk0NDRgb19fWgVhM2fTpDfHyYPGAASSoVni0tNIWE4BsfT9qx80QCAqTKWFiI2L2bhi1bKO7ulss92frYbFatosJeAwlyPb6FCxfyhz/8gblz57Jr164Trg94sjK6M4WtrvWFF16goqKC559/nn379tmFM6vVislkYsSIEVx99dW8/vrrLF++nCVLlgDn7zvwc1zQhFMoFFgsFoYNG8ZNN93E888/z1133QXAxo0bcXR0JDIyEo1Gg6enJyqVihdeeIHvv/+egwcPsm7dOoDz5j6oVCpG9SSGQ0JCeOWVVygqKuKKK67gscceIyEhgZdeeonuY+sSTxHr1q2jRq/HMTwc37o6unU6Cr//npDQUCZHR/Pc3//OnCVLCEpOJm3WLBS+vpJAtueqrweLBUVQEGNGj2bcsGG9n8BqlSMQjhngumXLFgYPHszGjRtZu3YtL7zwwgm7sA8cOEBtbS1bt24lOTmZFStWnPbrg6PhhNVqxWKx8OSTT5KTk8O3337LsGHD+Oqrr+xd/EqlEo1Gg5eXFzfccAO7d+9m8ODBWK1WhBCn/B3IysqyX+9ZWdLqfJvYU4Verxf/+te/hBBC6HQ6cd9994m33npLtLa2CiGEmD17tt29fOONN8TixYvFH//4RzF69Gixb9++Xuc61gU5FUybNk0EBQWJUaNGiQ8++OC3vxghRFtbm3j66adFcHCwCAoKEk899ZT9tZwKDh48KLKysuQ/BoNo+uwz8e0XX4g9e/bIx9rbheXgQbH6m29ERkbG0QN37RJi3Tohmpp6n7CyUogDB4TIzu7tUgr53i9evFg4OjqK+fPnC51O94vX9s9//lN89NFHQggh9u7dK+68885Tfl02nOgzuuOOO8SLL74ohJDv3w033CCee+65k16PucdlPlVs375dXHHFFWLo0KH278yZuv8nwwVPuBO98dnZ2eKBBx4QP/30kxBCiNLSUhEUFCS6urrExx9/LFJTU8VVV10lhBDi9ddfF2+++aYQQoiamppzcs179+4V48aNExMmTBDXXHONMBqNJ923q6tLLF26VERFRQkfHx/x0EMPifqffeFPCVar6NqzR1iPIa3FbBab3nxT7PvhByGqqoTIyRGisVGIjg4Zl7W19exokdtKSo477YEDB0RaWpoICQkR3377rT2u/iU888wz4uuvvxZCCFFQUCCuu+660389QoicnBxx//33iw8++EDU1dWJTz/9VDz55JOitLRUCCHEJ598IgICAsQ333xzHLlO5Tp/jqamJtHR0SGefPJJMXv27DO65l/DBe1SwtHx6eKYdGFSUhJ/+9vfGDduHACLFi1i2rRpODs7U1RUxPz58xk4cCBDhgzhmWeewdRTXLtw4UJmzZrFO++8g16vt59P9Mz26CucznJIzs7O3HfffeTm5vLcc8/x7bffEhsby913301ZWRn79u076bG9oFDgPGwYiqYmWR2CfO/Sp0xhiKurfCwhAbOHB6s2bGBfd7esFNm5U86HjI+Xg1h7YDabee655xg7dixxcXEcPHiQmTNnnpKMfqZldMe6cJs2beLqq69m9OjRfPXVVzz//PM4ODhgMBh47bXXMBqN1NbWMnnyZAYMGHCcy3iqcv+xz+nt7Y2rqyuLFi1CrVYzYcIEPvroo1M6zynjrND4LOPnd693331XNDQ0CCGEuOaaa8SXX34phBCiu7tb/Oc//xE7d+4U+/fvF1qtVixbtkzcdNNNYsOGDWLDhg19qmSeCEuWLBFfffXVKe9vMpnEZ599JpKTk4VSqRReXl4iJyfn9J60vl6I7duFyMsTwuYSmUxCHD4sTHV14ttvvxWZq1dL91GvP+7wgoICMW7cOOHj4yM++uij07YW+/fvF/PmzRNCCPH000+Lzz777Bf3P/b8tbW1wmKxiE8++USsXLlSFBQUiPj4ePH+++8LIaTFXbhwoUhPTxfz5s0TXV1dp3Vtx+JY76nNZu2FEC0tLSI2NlZER0eLsrKyMz7/ifC7JNwvYdmyZWLKlCni+++/7/WG3nnnneLhhx+2/79+/XqhUqnEggULxN13393LV7dYLKcd550IZWVlYsyYMb/oUp4IOp1O+Pr6ihkzZohRo0YJR0dHce2114pDhw6d+pdfrxfi0CEhfn5D2b1biP37pVv5M1itVvHmm28Kd3d3MXXqVFFeXn5a130s7r//fjFu3Dgxd+5cYTAYTrrfsa9n8+bNYvjw4eKHH34Qn3/+uQgPDxfjx4+3p4Ha2trs8dqx1/ZbPqu6ujpx/fXXi2eeecb+WENDgz0G7Wv81xFOCOnbjx07Vtx+++2ira1NHDlyRIwaNUrU1dUJIWQwfdVVV4mbb75ZNDQ0iHfffVcIIT/w4uLiPrmGtrY2MWHCBJGXl3dGx9uu1Wq1ivXr14spU6YIjUYjZs6cKTIyMk6deAUFUhApLpZx2kkEhqqqKnHZZZcJNzc38corr/TJDeeXcOz5DQaD2Lt3r/Dx8RHfffedEEKIrVu3ihtuuEEsX75cCCFEfn6+GD9+vFizZs1Jz3Oqz2fDf/7zHzFjxgy79TwRfulmcSb4ryLcz7+E1dXVQggh/vSnP4mFCxfaH9+zZ48YPny46Oi5y1dVVYkHHnhALFmyRKSkpIj77rvvOOXrdNQqs9ksLr/8crFu3bozfSnHwWq1ioyMDDFz5kyh0WjElClTxPr160+NeE1NQnR3n/S8y5YtE/7+/mLkyJGn777+RmRkZIihQ4eKwsJCMWLECHHHHXcIIYRob28X//nPf8SYMWPEjTfeKOLj48Vrr73WZ89bX18vnnnmmaNKr/htlvJU8bsg3OnGED9XrPbt29fLYtx2221iyZIlQgjpmjz00EMiICBArFy5UgghxM0332z/II51B2tqasSKFSt+VW7+7LPPhLe3t0hPTxfp6eli2bJlp3Tdhw8fFmPGjBETJkwQl1122QnlbqvVKjIzM8W1114rHB0dxdixY8W333572l8Wk8kkWlpaxHXXXSecnJzEX/7ylz6XwH+On3+Ob731lrjooovE5s2bhRBCHDp0SAwfPlxs3brVvk9lZaXYsmWLKDlGQT0TBdJsNouFCxeKt956SwjR+wZ6Juc7U1zQhDsbb0RpaalISkoStbW1Qggh8vLyxD333CPeeustceWVV4qUlBSRmpoqDh8+LGpqasTDDz8srr76anH48GHx5ptvigULFvT68PsSx5L7L3/5i/j3v//9i/sXFBSIW265Rbi4uIghQ4aIzz777JRI09bWJvz8/IS7u7tISEgQu3fv/s3X/ms49rpsr/PLL78UarVa5Obm2re9++67YurUqSe8gfxWC7Rq1SoxatQoe/x3Lizaz3FBE+7TTz8VDz30UJ+f10Y2IaSwkZKSYv9/z549YtmyZWLFihXi/vvvF/PnzxfffPONmD17trjuuuvE119/fdYtgRBCPPjgg6dMhIqKCrFo0SLh7u4ukpKSxDvvvHPS2KOzs1PcfvvtQqPRiMjISOHs7Hw0WX4WkJeX1+v9fvLJJ8VDDz0kduzYIUwmk3jooYfEnDlz7Nv1er2YPXu2eOGFF/rk+X/ujTz++OPiD3/4g/3/c2ndhLiACafX68WMGTPs1SUmk6lP3pyfn6Ozs1Pccsst4qGHHuolAb/44oti6dKloqqqSgghxNy5c8XMmTNFUVHRCc9bXFwsCgoKfvP1/fjjj2LQoEFi+PDhounn1SC/gvr6evHQQw8JHx8fERUVJZYuXdpLNs/IyBCJiYkiKirKHl9mZmaetRvI66+/LiZNmiQKCgqETqcT8+bNE/fff7/49NNPRWpqqli9erVobGwU06ZNs7t6QojTft2/hra2NvvnqNPpxPTp08Wrr74qhDj3Vu6CJdzKlStFenr6Cbf19V2ptrZW3HPPPSIxMVEsX75cdHV1iQceeED885//FELIAP7KK68UH3/8ca8PyPb3gQMHxMSJE4Wzs3OfXdPzzz8vnnvuuTM6trW1VTz11FMiKChIBAcHiyeffFI8+OCDwsnJSdx4442nVUJ2ptDr9eLKK68UP/74oyguLhabN28WDz74oNDpdOLmm28Ws2fPtue+tm7dKiIiIo5TiE+3NOtE+Pzzz0V8fLyYMWOGuOOOO0RDQ4NdNLPF6X3xPKeKC5JwRqNRXHPNNeKjjz4S5eXl4m9/+5tYunRpr30sFkufW7y2tjZ7y8eCBQvsMdTf//53sXDhwl6xhg1NTU1i3LhxYubMmeK2224Ter3+jK9Lf0wS+q233rLfhc8UnZ2dYunSpcLb21s4OTmJFStWnBMXynYjev755+2pjIaGBjF16lQRGhoqXn/9dfu+2dnZQgghVq9e/Zu/+D9/bWVlZWL69OmisLBQdHR0iPnz54ulS5eK5uZm8dJLL4mLL774Nz3fmeCCJNzWrVuFu7u7mDVrlrj66qvFRx99JHbt2iUyMzPFe++9Z6+ls+G3ugVWq/W4D3vVqlUiJSVF3H777cLT01O888479mDf5oJt27ZN3HPPPeLPf/6zKCwsFFdccYU91WCDXq8/qRv6c6xatUpMmDBBTJw4UVxzzTV9VgXT0tLSS/4+Fu3t7WLEiBHC1dX1tPv9fg7be2j7PB555BExePBgcfvttwshZPz0wAMP2N3cu+66S8ybN6/XjeZMbwgnImt1dbUYOnSo3XLu3r1bzJgxQ5SVlQmDwSBmz559TgSjY3HBEc5isYgFCxaI+Ph48ac//UkIIb+0n332mUhLSxP33XefSE5OFs8++2wvolmt1uOI1xdE/Oabb8Rf/vIXcejQoV7bjEajGDRokHj55ZeFEEL84x//6BX8Nzc3i1deeUXMnDlTXH755WLKlCn2JO6FBJPJJOrr68+owVYI+R4dGwO2tbUd5wmEhISItWvXirq6OvHYY4+JUaNGiQkTJohrr71WtLS0/Kbr/zlBlyxZIn744Qe7Evnggw+Kzz77zP5duOKKK8Q333wjhOj7pPap4IIj3IYNG0RiYqJobGwUV1xxhejs7BRff/21eOSRR+xVBq2trWLt2rXCbDaLBx98UBw4cOC48/xWIeDnd8xjz7d9+3axZMkScdtttwkh5A3h8ccft7fu5OXliSuvvFLMmTPHnmP66aefxJQpU8Szzz57TmOGU8WZEK6tra1XPLh27VqRlJQkRo4cKT755BN74cGKFStEdHS06O5Jvh86dEjs2rXLftyZvh/H3lA7OzvFgw8+KC655BLxwAMPiJtuukmYTCaxYsUKsXDhQrFkyRKxatUqMWTIELsbez5wwXULhIeH89hjj+Hj44Naraa4uBghBFarlUmTJiGEwMPDg4svvhiTyUR8fDx33nknCxYssHcWNzQ0cPHFF1NWVnbG12GrPrdVkx+7ks6yZct45ZVX7CPBHR0dUalUtLa2otfr+eCDD/D09OTtt9+2T/adMmUKDz30EG5ubqhUqj7tTjhf+Pvf/25/fU8//TRvvvkmP/74I3/961/ZsmULW7duxWg0ctVVVzFq1CguueQSQA5kHTFiBHDm80WOHZG4dOlSnnnmGdzd3fnhhx944IEHcHV15R//+AdXXXUVCxcupKGhgU8//ZSXXnqJpGOH255jXHCEi4mJ4Y9//CMAycnJrFmzhqqqKtra2nB2dra3XTQ2NvLBBx/g4eHB9u3bsVgs9inIX331FVqt9uhaZr8BJ5p7+eqrr/LRRx/x5ZdfMnToUL788ku+/fZbhg8fTlZWFv/f3tmHNPm1cfxCLRGn4ls2X0B8yUjTVEzJ6YSWlUpCJEmYc6SkhKmQpoVJKcICzSKp0Or3h2HRC4S/QlAQxzIpYmS2lhn5VhRkFjZkm9v3+WPP7mdT6ymbe7H788/cdp975z7H731f51zXuc709DTt27ePPDw8aHZ2liknEAioqKjoj+tkDpqbmyktLY2am5uXfY76+npau3YtNTc3k6+vL0mlUgoMDKQdO3bQli1b6NmzZ0wC1hs3bpBQKCQi06VWy92Lz8HBgd69e0dVVVU0NDREExMT1NPTQxMTE+Tj40MikYiePHlCXV1dFBsbSy0tLdTZ2Umpqakrs5L7V7Has/UHLLTJVSoVhoaGkJiYiNraWnz69Alv3rxBYWEhCgoKIBKJEBERAU9PT6jVaiiVSvD5fGZxqrnNt4Xjwu7ubvB4PHh4eEChUEAikSAlJYWZ7bx+/Tp4PB7u3Lnz0+u0Nr9rUhra4cWLF/D19cXQ0BAOHjzIJO35+vUrKisrUVtbi6mpKbPU0bjt1Wo18vPzERMTw7zPy8vD1atXMTc3B61Wi0uXLkEkEpn4cK0RXWKMzQnOwMKGGRsbQ2VlJcRiMV6/fm2yIjc3NxfHjh0DALS3tyMjI+O3zr0cFgp5ZGQE09PTUCgUi36/oqICmZmZmJubY2I6bYnlppAwtGN9fT22bduGt2/fIjk5GVKpFIB+Fvdf42xhy+Tu3bvMDLFx38lkMnC5XPT19QHQjyH379/PvFcqlTY3XraLzMvG9rpOp6MPHz5Qeno6ZWdnk7+/P508eZIGBwcpJCSE9uzZQ0eOHKHs7Gwm2SeRPrHN+Pi4fisoM6LVasnBwYExdVUqFVVWVpJEIiGhUEhOTk706tUrEggE5OrqShcvXqSEhAQqLS1d8Xz7liQlJYXS09MpLi6Ojh8/TsPDw2Y5r06no9zcXPL19aXW1tZF37e1tVFnZyf19PSQo6Mj1dTUkEajofr6enL572Yjv5sSf0WxtuJ/FZ1OZ2KGff/+HdeuXUNaWhry8vIA6Nc37dy5c8nyAoEAkZGRyMnJYSJIDKzEXXB0dBTl5eUoKyvDwMAAlEolTp06hdLSUov7flYSQ59MTk7C2dkZfX19zLT7n5jNxmXHx8eRmpqK+/fvAzDtL41Gg8LCQiZRkVqtXuQLtSXsRnDGGPtPVCoVRkZGoFQqweFwUFRUtOiYf/75B05OTujt7cXg4CDc3d1RV1eH06dPm93E02g0i1wScrkcNTU1EIlEZv0tW8FwvYcOHTKJiVwuS/nHysvLweFwmKAH42O6u7vh7u6OwcFBk4SxtohdCs7Awo45ceLEksdt2rQJPB4PANDX1wcnJyfw+XzEx8fDycmJGXMY0Gq1f9xhGo2GGW80NTWhoqKCicq31X8GW0KpVCIpKQlCoZBxpMfExCA8PJw5xng8Z8hpY+vYteB+hZs3b5qELSUnJyMtLY3pxOjoaBMT0ziZjDlntKwR1WBPGN+Erly5Ag6Hg/T0dPB4PLi5uWFmZgbDw8Pw8vJiQsUWtqm1ZyB/BRsZSa4cHR0dFBgYSFFRUdTT00Pj4+OUk5NDoaGhJJFIaHp6mtknrqSkhGJjYyk4OJgkEgkz0P78+TNVVVUR0fKz8f7pXnSrmfn5eSawYGpqigBQbW0tNTQ0kEKhIKVSSZmZmRQZGUllZWXU1tZGL1++XNSmNjMx8jOsrXhLYIjX27p1K2JiYjA2NgatVov8/HyEhIQA0KdV8PLywrlz55CRkYENGzYw47vi4uJFS29Ys9C8tLW1wdvbGwcOHACgDzT29PREY2Mj5HI5iIjJrC0Wi61Z1T9iVQvOYGIYXh88eIBbt24B0PuIgoKC0NDQAJlMhoSEBKZDZ2dnsWbNGsjlcszNzcHPzw9Hjx5lvjOGFd7ymDNKatTb2wsOh8P4UgF9/3h6ejJmo4uLC6KiokzK2YMJuZBVLbifceHCBfj5+QHQT6R4eXlhYGAAAFBdXY3NmzdjZmYG1dXV8PDwYMoZUn8XFhZao9p2T11dHfN3f38/s1+At7c387lGo8Hz588RFxcHf39/uLm5IScnxwq1NT9/reAAMMvuHz58CB8fH+bzwMBACIVCjI6OIigoCEKhEMD/nma3b99GfHw8uFyuyZKcyclJFBQUoL293XIXYUeMjY3BxcUFBQUFAICSkhLs3r0bUqkUoaGhjPUB6Nuyv7/fJFwMsP/Jp79WcMbmyOzsLLZv3w5HR0cEBAQgLCwMGo0GTU1NcHNzW7IMoPf/LHzSVVVVISwsbMnV4Sz6dYOurq6QyWSIj49HdXU1ZDIZUlJSkJSUBKlUiq6uLnC5XGanHAP2aEIu5K8V3FIoFAp0dXUxrgEul8uYMsZ3VmMfmzFarRZnz57F3r17LVNhO4XP52P9+vXYuHEjk4Py6dOnEAgEWLduHZydnZm1hqsNu4ilXGl0Oh3Nz8+bTDM/evSIsrKy6P3798xOqkuVM3YdzM/PU1FRER0+fJiysrJsK4bPhpiYmKDExET6+PEjhYeHk1KpJD6fTxEREZSRkUEBAQHk7+9PRDYWB2kGWMH9BENnG/uJfnTcmTNn6Pz588Tlckkul1uwlvZJR0cHFRcXU2trK6lUKrp37x4FBwfT5cuXiYj+b5vbK6zgfoBarf5tZ3VLSwv19vZSY2MjRUdHr1DNVg+7du2ix48f07dv36xdFYvBCo7Fanz58oXEYjGJxeJV+0RbCCs4M7Laxhss5ocVHAuLBWFvxywsFoQVHAuLBWEFx8JiQVjBsbBYEFZwLCwWhBUcC4sFYQXHwmJBWMGxsFiQ/wAeQyZgUiO5TwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 220x220 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "%config InlineBackend.print_figure_kwargs = {'bbox_inches':None}\n",
    "\n",
    "import numpy as np\n",
    "from scipy.optimize import curve_fit\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "# Define the modified sigmoid function in 3D with upper and lower plateaus\n",
    "def sigmoid_3d(xy, a, b, c, lower_plateau, upper_plateau):\n",
    "    x, y = xy\n",
    "    return lower_plateau + (upper_plateau - lower_plateau) / (1 + np.exp(-(a*x + b*y + c)))\n",
    "\n",
    "# Define data\n",
    "x_data = predicted_stability_vs_binding['folding_ddG'].to_numpy()\n",
    "y_data = predicted_stability_vs_binding['binding_ddG'].to_numpy()\n",
    "z_data = predicted_stability_vs_binding['binding_fitness'].to_numpy()\n",
    "\n",
    "# Initial guess for parameters\n",
    "lower_plateau=min(z_data)\n",
    "upper_plateau=max(z_data)\n",
    "initial_guess = (1, 1, 1, lower_plateau, upper_plateau)\n",
    "\n",
    "# Fit the modified 3D sigmoid surface to the data\n",
    "params, covariance = curve_fit(sigmoid_3d, (x_data, y_data), z_data, p0=initial_guess)\n",
    "\n",
    "# Extract the fitted parameters\n",
    "a, b, c, lower_plateau_fit, upper_plateau_fit = params\n",
    "\n",
    "# Create a meshgrid within the range [-2, 2] for plotting the surface\n",
    "x_mesh, y_mesh = np.meshgrid(np.linspace(min(x_data), max(x_data), 100), np.linspace(min(y_data), max(y_data), 100))\n",
    "z_fit = sigmoid_3d((x_mesh, y_mesh), a, b, c, lower_plateau_fit, upper_plateau_fit)\n",
    "\n",
    "# Calculate the minimum and maximum values of the fitted surface\n",
    "z_min, z_max = np.min(z_fit), np.max(z_fit)\n",
    "x_min, x_max = np.min(x_mesh), np.max(x_mesh)\n",
    "y_min, y_max = np.min(y_mesh), np.max(y_mesh)\n",
    "\n",
    "# Plot the original data points and the fitted wireframe surface\n",
    "fig = plt.figure(figsize=(2.2,2.2))\n",
    "ax = fig.add_subplot(111, projection='3d')\n",
    "ax.scatter(x_data, y_data, z_data, color='black', label='Original Data', alpha=0.5, s=0.1, rasterized=True)\n",
    "surf = ax.plot_wireframe(x_mesh, y_mesh, z_fit, color='red', label='Fitted Surface', rstride=5, cstride=5, linewidth=0.1)\n",
    "\n",
    "# Set xlim, ylim, and zlim to touch the axes\n",
    "ax.set_xlim(x_min, x_max)\n",
    "ax.set_ylim(y_min, y_max)\n",
    "ax.set_zlim(z_min, z_max)\n",
    "\n",
    "ax.set_xlabel(r'Inferred ${ΔG_{f}}$'+ '\\n'+ r'kcal mol$^-$$^1$', fontsize=7, labelpad=-8)\n",
    "ax.set_ylabel(r'Inferred ${ΔG_{b}}$'+ '\\n'+ r'kcal mol$^-$$^1$', fontsize=7, labelpad=-8)\n",
    "ax.set_zlabel('Binding fitness\\n(observed)', fontsize=7, labelpad=-8)\n",
    "ax.tick_params(axis=\"y\", pad=-6)\n",
    "ax.tick_params(axis=\"z\", pad=-4)\n",
    "ax.tick_params(axis=\"x\", pad=-6)\n",
    "ax.set_xticks(np.arange(-1,4,1),labels=np.arange(-1,4,1),fontsize=6)\n",
    "ax.set_yticks(np.arange(-1,3,1),labels=np.arange(-1,3,1),fontsize=6)\n",
    "ax.set_zticks(np.arange(-1,0.5,0.5),labels=np.arange(-1,0.5,0.5),fontsize=6)\n",
    "\n",
    "ax.set_rasterization_zorder(0)\n",
    "\n",
    "ax.grid(False)\n",
    "ax.xaxis.pane.set_edgecolor('gray')\n",
    "ax.yaxis.pane.set_edgecolor('gray')\n",
    "ax.zaxis.pane.set_edgecolor('gray')\n",
    "ax.xaxis.pane.fill = False\n",
    "ax.yaxis.pane.fill = False\n",
    "ax.zaxis.pane.fill = False\n",
    "ax = fig.add_axes(MyAxes3D(ax, 'lr'))\n",
    "ax.view_init(elev=30, azim=310)\n",
    "\n",
    "# Unpack the meshgrid values before passing to legend\n",
    "#ax.legend(handles=[plt.Line2D([0], [0], marker='o', color='w', markerfacecolor='red', markersize=5),\n",
    "#                   plt.Line2D([0], [0], color='blue', linewidth=1, label='Fitted Surface')],\n",
    "#          labels=['Original Data', 'Fitted Surface'])\n",
    "\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storder3statemodel_folding_n_bidning_phenotypevsaddG.svg', dpi=300, transparent=True)#, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11ded8d7-b5bd-422e-ab71-370f88d0fb38",
   "metadata": {},
   "source": [
    "## Panel b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "f566d132-ed4d-49d7-b81c-a07bce4d8dba",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 140x160 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "%config InlineBackend.print_figure_kwargs = {'bbox_inches':'tight'}\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "\n",
    "overall_full_model_performance = pd.DataFrame()\n",
    "\n",
    "y = predicted_binding_scaled.loc[predicted_binding_scaled['WT']!=True].loc[predicted_binding_scaled['predicted_binding_fitness'].notnull()][['binding_fitness']].to_numpy()\n",
    "x = predicted_binding_scaled.loc[predicted_binding_scaled['WT']!=True].loc[predicted_binding_scaled['predicted_binding_fitness'].notnull()][['predicted_binding_fitness']].to_numpy()\n",
    "model = LinearRegression()\n",
    "model.fit(x,y)\n",
    "y_pred = model.predict(x)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(1.4,1.6))\n",
    "hbin=ax.hexbin(x=predicted_binding_scaled['predicted_binding_fitness'].loc[predicted_binding_scaled['WT']!=True],y=predicted_binding_scaled['binding_fitness'].loc[predicted_binding_scaled['WT']!=True],bins='log',gridsize=75,rasterized=True,linewidths=0,vmax=1000)\n",
    "ax.axline((0.1, 0.1), slope=1, linestyle = \"dashed\", color = 'black',lw=0.5)\n",
    "\n",
    "\n",
    "plt.axvline(0, ls='--', color='black', lw=0.5)\n",
    "plt.axhline(0, ls='--', color='black', lw=0.5)\n",
    "plt.title('\\nR$^2$ = '+str(round(r2_score(y,y_pred),2)),fontsize=7, color='r', loc='right')\n",
    "overall_full_model_performance.at['1','All']=round(r2_score(y,y_pred),2)\n",
    "plt.xlabel('binding fitness\\n(predicted)', fontsize=7)\n",
    "plt.xticks(fontsize=6)\n",
    "plt.ylabel('binding fitness\\n(observed)', fontsize=7)\n",
    "plt.yticks(fontsize=6)\n",
    "ax.set_xticks(np.arange(-1.5,0.5,0.5))\n",
    "ax.set_yticks(np.arange(-1.5,0.5,0.5))\n",
    "ax.set_ylim(-1.3,0.2)\n",
    "ax.set_xlim(-1.05,0.2)\n",
    "ax.spines[['top','right']].set_visible(False)\n",
    "ax.set_rasterization_zorder(0)\n",
    "fmt = '%1.f'\n",
    "cax=fig.add_axes([0.15,0.82,0.3,0.05])\n",
    "cb=fig.colorbar(hbin, ax=ax, shrink=0.7,format=fmt,location='top',cax=cax)\n",
    "cb.outline.set_visible(False)\n",
    "cb.ax.tick_params(axis='both', which='both',direction='in', color='white',width=0.25)\n",
    "cb.ax.xaxis.set_tick_params(pad=0)\n",
    "#cb.ax.set_title('Count', fontsize=7)\n",
    "#cb.set_label('count', rotation=0, fontsize=7, labelpad=2)\n",
    "#cb.ax.xaxis.set_label_position('right')\n",
    "cb.set_ticks([1,10,100,1000])\n",
    "cb.set_ticklabels([1,10,100,1000], fontsize=6, rotation=45)\n",
    "ax1 = cb.ax\n",
    "ax1.text(1200,0.1,'count',rotation=0, fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storder3statemodel_binding_performance.svg', dpi=300, transparent=True, bbox_inches='tight')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd01bf2a-7c0a-4701-ad3c-42559b5175a4",
   "metadata": {},
   "source": [
    "## Panel c"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "bba23275-b7a6-459d-b4e6-ed168e873387",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAKoAAADVCAYAAADHAmuWAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAlYElEQVR4nO2deVhV1frHv2dAREVEMAcUFAcQ8IgMIuIFUbCcRQa1VHrKzLqaORQ/cySzrt0kNTMyb1qaCSqC4ixiKCrOCIqhoAmSA4iAoAfO3u/vDzo7jgxnAuTk+jzPedh7rfXu993nfFl7rb3XWltERAQGo4kjftEBMBiawITKMAiYUBkGARMqwyBgQmUYBEyoDIOACZVhEDChMgwCJlSGQcCEWgcffvghvv/+e2Hfzs4Oq1atEvY9PDwQHh4OZ2dnODs7o1mzZpDJZHB2dsZ///vfeotjyZIlOHHiRLX0zZs3Y/78+fXmp0lDjFr59ddfKTQ0lIiIHj16RP369aPAwEAiInr27BlZWFiQQqEQytvY2FBJSYna41a10YdNmzbRvHnz6uVYTR1Wo9aBp6cnzpw5AwA4c+YMxowZg7y8PADAxYsX4ezsDIlEotGxli1bhhkzZsDPzw9z5szBjRs34O/vD1dXVwwbNgx//vknAGD16tWws7ODTCbDe++9BwB48803ER8fDwCIi4tDr1694OPjg5MnTwrHHzx4MNLT0wEA6enpGDx4MADgyZMnmDp1Ktzd3eHm5obk5GT9v5gXgPRFB9CUsbGxQUlJCQoLC3HmzBl4enoiIyMDd+7cEfa1IS0tDYmJiWjWrBn8/f2xceNG2NjYYMeOHQgPD0dkZCQ+/fRT5OTkoGXLligqKlKxf/r0KWbNmoXk5GR06NABvr6+GDBgQJ0+P/vsMwQEBODnn39Gbm4uRo4cidTUVK2/ixcNE6oaPDw8kJKSgpSUFMyePRvXrl3D6dOncebMGYSGhmp1rLFjx6JZs2YoKSlBcnIyxo4dCwDgOA5du3YFAPTv3x+TJ09GcHAwxo0bp2L/+++/w97eHl26dAEAhISE4M6dO3X6PHLkCPbv34/w8HAAQEFBAcrLy9GsWTOtYn/RMKGqYcCAATh9+jQKCwvRtm1beHh4YOfOnUhJScH69eu1OlaLFi0AAESEDh064PLly9XK7Nu3D8ePH0dMTAwiIiJw/vx5IY+IIBKJajy2VCoFz/MAALlcrmITHx8Pa2trrWJtarA2qho8PT2xZcsW2NnZAQBcXV2xf/9+NG/eHBYWFjods3Xr1mjXrh32798PAKioqEBGRgZ4nkdOTg6GDh2KiIgI3L59GxzHCXb29va4fv06cnNzoVAosGPHDiHPxsZGEH5MTIyQ7ufnh2+//VbYN8TLPsCEqhY3Nzfk5ubCw8MDANC8eXOYmpqqbRuq45dffkFERAT69u0LZ2dnpKSkgOM4vPHGG5DJZHBzc0N4eLhKZ83ExARr167F0KFDMWTIEPTr10/Imzt3Lr788ksMGjQI5eXlQvqSJUvw4MED9OnTBw4ODti4caNecb8oRERshD+j6cNqVIZBwITKMAiYUBkGARMqwyBgQmUYBEyojHqjIW8gMaHWwIMHD16Yb+XTJUOB53kkJCQAQK1PzeoDJtQq8DyPqVOnIi4uTtjXxvabb77ByZMnceXKFZ18z5kzB9nZ2QAatnaqT7Zu3YpXX30ViYmJQlpDxM6E+hc8z2PKlCm4c+cOjhw5grKyMojFmn89b731Fq5fv47MzEyEh4fj0KFDWvkODAyEvb09unfvDp7nUVZWprE9EWH79u0oLy/XuUau+jRLG3r37g13d3d89NFH2LVrF4C/a9Z6FeyLGgjbUPA8r5PdmjVraNGiRUREtGzZMgoLCyOO4zSyvX37NgUEBJBcLiciotWrV9PAgQPpxIkTGtnv3buXTExM6NKlSzRjxgx69913af78+ZSUlKTWluM4Gjt2LJmZmVFGRgYRafcdcBxHYWFhNHPmTEpMTKS7d+9qbKtk586d9Ntvv5GXlxft3r2bjh8/rnUc6vjH1ai6tpPGjBmD5cuXAwCCg4PRvHlzPHv2DID6msHGxgYDBgzAjz/+CJ7n0bdvX7i6uuLYsWMoKytTaz9q1Cj88MMPcHFxgUwmQ1hYGAYPHoxt27ahoKCgVnsiQmhoKLy9vbFhwwYsXboUjx490uo7eP/990FECAoKwoEDB7Bp0ybcunWr1vLKJkpOTg6ICOXl5Th+/DhatmyJzZs3Y+bMmVi3bh2A+m2z/iOG+fE8j/feew/GxsYICgqCt7d3nUPilBARTp06BS8vL2E8KAB069YNv//+O9atW4ePP/64xuPwPI+YmBgEBQUBAAYNGoTExEQEBASguLgYH330ETIzM9G8efNa7efOnQuJRILAwEC88cYbcHBwEAaaWFpa4uDBg5BKpbWex/Xr1+Hn5yeMi83IyEBOTg7atm0Lnuc1aro4ODhg0KBBcHFxQYsWLTBnzhx07twZ3bp1q/Yd8jyPcePG4bXXXkOXLl3w9OlTmJiYYPbs2UhISEBaWhp8fX3x4MEDyOVyGBkZadV8qguDr1GJCMHBwbC0tISnpyd27twJjuM0Eqmvry/Gjx+PH374QUhXKBQwMTHB6tWrcerUqRrHjAKAr68vQkJCBNuBAwfi448/xsqVK3HgwAGYm5sjJSWl2ih9pe/AwEC0bt0ajo6O+Omnn5Cfn68yGurIkSO4du2aUKs/b//gwQN069ZNECnP8zAyMhKG9NUlECJCUlISOI6DmZkZIiIicO3aNVRUVKB169aIj4/HvXv3qn2HGzZsAAD069cPs2fPRmhoKDZt2oS8vDxcvHgRlpaW2LJlCw4dOgRjY+N6E6kyaIOF4zhatWoVrV+/Xkjz9/en//u//6OoqCh6+PBhrbbXrl2j//3vf1RQUEDe3t70ww8/CHlyuZzKyspo/fr1VFBQUM3nsWPH6PDhw5SXl0ejRo1S8U9ElJOTQx4eHnT16tUaY169ejWtWrVKSPvkk08oIiKCiIgqKiro9OnT1LdvX7p27VqN9sOHD6fp06eTh4dHtTblpEmTaNu2bbWeNxHRwoULycHBgXbv3k1ERJGRkbRw4UIaP3485eTkUGRkJF26dEnF5sKFCzR//nz6+eefycvLi9atW0cXLlygsLAwio6OVvmu62vyYlUMVqgcx9HkyZNp4cKFQlpcXByNGTOGfvrpJ5o9ezZdvny5Vnue56mwsJCIiLKysmjQoEG0YcMGIiIqLS0lIhI6R1V9Tpw4kRYtWiT8GLdu3aLXXnuNIiMjiaiyY0VEVFxcXGfMJSUlQmdj165d9PXXXxMRUW5uLuXl5dGDBw9qjHnx4sX0ySefEBHR2rVraenSpcTzPJWXlxMR0ZYtW2j9+vV1iiUzM5MmTZpE33zzDW3dulUl7/jx4+Ti4kK5ubkq6QqFgt566y2aNGmSiohv3LhBo0ePpvz8fCHGhsBgL/2BgYEoKCjAZ599BgDIzs7Gv/71L8TFxWHq1KkwMTFBSkpKNbulS5ciLi4OIpEIbdq0AcdxsLW1xZYtW7Bz5068/vrrWLx4MRQKRbV5RYGBgSgqKsLy5cshkUiQkZGBrl27YsOGDThy5AgmT56MZcuWobi4GKampjXGnJ+fj88++wytWrVCVlYWAMDa2hrm5ubYs2cPFi1aBFNTU7Rr166a/YoVK3D//n2sWLECAODu7o7S0lKIRCIYGRkBqJw6ExAQUOPsWOW5d+vWDU+fPoVYLEZ+fj62bduG5ORkPH78GOnp6fjll19gZWUFIsLDhw8hl8shkUgQERGBVq1aQaFQCMfMyMiAWCwW/DfYTf8GkX8jsHfvXpoyZQqdP3+eli1bRlOnThVuJ0VHR9PgwYMpKyurmt2BAwdowoQJdPjwYSKqrOWUtdHatWvJ2tqa0tPT1foMDw+nKVOmCLZr1qwha2vrGi/XddkrFApKSkqidu3a0ZAhQ+q0P3r0KE2cOJEOHTpERET3798X1h3Yt29ftWbK8+zfv59CQkLo0qVLdPz4cbp8+TKdPHmS+vXrRwEBAcRxnMotudTUVBoxYgStWLGC4uPjiYjo888/F7b37NlDvr6+lJaWVqff+sBghUpElJiYSM7OzuTp6SmkZWZm0qxZs4R7ijVx7NgxCg4OpgMHDghpJ0+epIULF9ZpV5tPhUJB27dvr1Nkddn/8ccfFBAQQL///rtae2XsSrG+8847FBkZSf3796dbt26ptT969CiFhITQtGnTaNWqVTRnzhwKCAigdevWUVlZWbXyDx48oIMHD5KHhwdt27aNli9fTp6ennT37l3KzMykO3fuqPVZHxi0UImIDh06RIGBgXT69Gkiqmwj1dQ+fJ6EhAQKDg6mCxcu0NGjR2nEiBGUl5enk099Y+Y4jh49eqSxfUJCAgUFBdGRI0do0qRJ5OvrS5mZmRrbnzhxgiZPnkydOnWiL7/8kogqO3F1cePGDVq7di19/vnn1KpVK/r+++819lcfGLxQiSprqWHDhtHRo0e1sjt+/Dg5OzuTTCbTqDaryWdCQoJWds/bHzlyRGf70aNHk7u7O12/fl0rW57n6dSpU+Tj4yM8PdOkE6TsoC1fvlyj2rs++cdM7jtx4gRsbW1hZWWlld2xY8fQuXNn9OrVq9F81pf9+fPnYWVlhY4dOzaKf9LgIUpD8Y8RKuOfjcHenmK8XDChMgwCJlSGQcCEyjAImFAZBsE/SqhyuRzLli1TWXaxseyZb918a8o/6vZUcXExzMzMUFRUhNatWzeqPfOtm29N+UfVqIx/LkyoDIPAoOZM8TyPvLw8mJqa1vgor7i4WOWvtuhjz3zXbEtEKCkpQadOnfSammJQbdTc3FzhRQsMwyInJwedO3fW2d6galTlqPlzvx1Bq1YtdTpGb9eBOvtPu3hefaE6MOJ17xmblBXo5ZuXGOls2/xxns62JWVP0St4Zo0zHrTBoISqvNy3atUSpq1aNbp/U1P9fBpxuoulhbj6bFRt0EuoFS308g3oP0WFdaYYBgETKsMgYEJlGARMqAyDgAmVYRAwoTIMghcq1FWrVtWZL5fLUVxcrPJhvJy8UKGqW5X5iy++gJmZmfBhT6VeXpr0pX/BggUoKioSPjk5OS86JMYLolGeTAUHB1d7MkFESEtLq9PO2NgYxsbGDRkaw0BoFKF+9dVXjeGG8Q9GL6GmpqYiKioKn3/+eZ3lbGxs9HHDYGgv1PT0dERFRWHHjh2wtLREQEBAQ8TFYKigkVAzMjIQHR2N6OhomJubIzg4GGVlZTh58mRDx8dgANBQqI6OjvDy8kJsbCx69uwJAFizZk2DBsZgVEWj21OHDx+Gg4MD/P39MXHiRMTExBjcOzsZho1GNaqfnx/8/PzAcRwSEhIQFRWFsrIyBAYGYty4cZgyZUpDx8l4ydGqMyWRSDBs2DAMGzYMCoUChw8fxo4dOxpdqIWidqgQ6za1IT/8HZ39GhXf0dkWAFq7+OtsW3oqRi/fohvpuvt2f01n2zKTJzrbVkXnJ1M8z6O8vLzBV8hgMAAthVpRUYH4+HhMmTIFHTt2xM6dOzF+/PiGio3BENDo0n/w4EFERUUhISEB3t7eCA4ORlJSErZu3drQ8TEYADSsUUeMGIGbN2/i8OHD2Lp1K8aOHVvjC7cYjIZCoxr15s2biIqKwqRJk2BsbIyQkBCVt7cxGA2NRjWqra0tFixYgEuXLuHnn39GaWkpzMzM4OrqKrzjnsFoSLTu9ffq1QuLFy9GWloatmzZAo7jGiIuBkMFvUZPOTg4YNmyZWrLrV+/vta8999/X58QGC8JjTIeddGiRejVqxcCAgLQsWNHaLoum1wuV7lPy+ZMvbw0ylSU+/fvY+nSpcjKykJsbCyKiorg7++P0NDQOu3YnCmGEq1r1NzcXPzxxx8qbVNvb+86bYyMjDB8+HAMHz4cSUlJmDNnDrKzs7F69eo67RYsWIC5c+cK+8XFxUysLylaCXXevHnYu3cvnJychPuoIpFIrVB/++03xMTE4M6dO/D09ER0dDS6d++u1h+bM8VQopVQ4+PjcfXqVRgZabeEoa+vL9zc3GBjY4Nz587h/Pm/1xmNjo7W6liMlxOthOrg4ICSkhK0bdtWKye3bt3SqjyD8TxaCbW0tBT29vbw9PRUuSSrqxXZ5D6Gvmgl1IULFzZUHAxGnWglVB8fH9y7dw/nzp2DSCSCu7s72rdv31CxMRgCWt1H3bZtGwYNGoR9+/Zhz549+Ne//oXt27c3VGwMhoBWNerKlStx7tw5mJubAwAKCwsxePBgTJw4sUGCYzCUaCVUnufRqsrbSFq1avVCZqNKRQpIRboNM2zpNUh3xzm/624L4LFI9zG8pZY99fJt0qqdzravOHro5bs+0EqokydPxsCBAxEYGAgAiImJwdSpUxskMAajKloJNSwsDH5+fsIKKd999x1cXV0bJDAGoyoaCVUul8PY2BhlZWXo3bs3evfuLeSVlZWhRQv9X5jFYNSFRkINDAxEfHw8HB0dVdY5JSKIRCJkZ2c3WIAMBqChUOPj4wGwR6GMF4dW91E9PT01SmMw6huNatT8/Hzcv38fRUVFyMjIEEboFxcXo7CwsEEDZDAADYW6b98+bN68GTk5OSpznExNTdWuNs1g1AcaCTU0NBShoaGIjY3FuHHjtHby5ZdfIigoCLa2tlrbMhiAlvdRU1NTceXKlWrpS5YsqdPO0tISs2bNQmFhIcaOHYvg4GCNRMsm9zGUaNWZsrS0hIWFBSwsLNCqVSscP35co3c/vfXWW9i3bx8OHDiAjh074sMPP8TAgQOxcuXKOu3Y5D6GEhFpOne5BioqKjBy5EgcPnxYY5tHjx4hNjYWmzdvBsdxSE5OrrVsTTVqly5dcPZSBlqZ6rY+arfbCTrZAQCe6FejP7N20Nm2tKXuz+oBwOTZY51tLfrWPSdOE4qKitC6dWud7fWa13/v3j2NbvY/fPgQu3btQkxMDMrLyxEQEIBt27ahc+fOddqxyX0MJVoJtV27dipPptq0aYMVK1aotQsMDERQUBA2bdoEKysr7aNkvPRoJdSHDx/q5CQpKUknOwZDiVadKeXwPnVpDEZ9o1GNqlAoUF5ejqysLDx9+lTlydS1a9caNEAGA9BQqN9++y1Wr16NvLw8ODo6CkJt3bo1W42P0ShoJNTZs2dj9uzZWL9+PRMm44WgkVCTk5Ph5eWFrl27Yv/+/dXyR4wYUe+BMRhV0Uiohw8fhpeXF3bs2FEtTyQSNbpQ+/frrb5QLcRLe+lsO0qRqbOtvhT+59962f95VveJiU+Sd+lsW1xahk7D9H9hnkZCDQ8PBwDMnz8fjo6OKnlXr17VOwgGQx1a3Z6q6VWS7D2ojMZAoxr1/PnzSElJQX5+vsp6/MXFxexlE4xGQSOhlpaWIj8/HxUVFSpPp0xNTbFz584GC47BUKKRUH18fODj44O3335bGEgil8tRUFCATp06NWiADAagZRt11qxZKCkpQVFREXr37o2hQ4eqHTTNYNQHWgn19u3bwuV+0qRJuHbtGmJjYxsoNAbjb7QSqlwuR1ZWFrZt24YxY8aoDPljMBoSrYS6ZMkSBAQEoFevXvDw8EB2djZ69Oih1u7EiRPCdklJicb+5HI5iouLVT6MlxOthDpx4kRcuXIF3333HYDKl/nGxMSotVu6dKmwPXbsWI39sTlTDCVaCfXmzZsYOXIkOnXqhE6dOmH06NHIyspqqNiwYMECFBUVCR9NJhIy/ploNcJ/0qRJ+OSTTxAQEAAAiI2NxcSJE3Hu3Lk67dLT0xESEgIiUtkWiUR1vlGFzZliKNFKqFKpVBApAIwbN07tlGcAaoXMYKhDI6EqR/F7eHjgww8/xMSJEyESiRAVFYUhQ4aotWfvmWLoi0ZC/fe/VYeYpaamCtvsFhWjMdBIqImJiQ0dB4NRJ1q1UT/99NMa09ljVEZDo5VQLSwshG25XI74+HiNXmfOYOiLVkJ9vq06a9YsjBw5sl4DYjBqolHWnqpv9Flw61nUlzr7Lemq+1wtAHjW0lJnW/M+A/XyrQ+5X0TobFvS/Em9xKDT2lPKm/Warj3FYOhLo6w9xWDoi0bP+v/44w8UFRUJ+0ePHsUHH3yAiIgIlfVLGYyGQiOhBgcHo7S0FEDlRL+QkBBYW1sjLS0N7777boMGyGAAGl76nz17JsyN2rp1K6ZPn4758+eDiCCTyRo0QAYD0LBGJSJhYbQjR47A398fAHt8ymg8NKpRQ0JC4O/vD3Nzc0gkEvj6+gIAsrOzYarjWvoMhjZoJNTFixfDz88P9+7dg7+/P8TiyopYoVBg3bp1DRoggwFocXuqpnee9uql+4JjDIY26PVkqqFhL0RjKGkUoQYHB1freGkyFeWLL74QVhJkvNw0ilC/+uornewWLFiAuXPnCvvKF6IxXj4aRai6TkVhk/sYSrSaLs1gvCiYUBkGARMqwyBgQmUYBEyoDIOACZVhEDTpJ1PPU/UdrLryrOyZzraK0jKdbQFATqV62b8oSp7oPu/pyZPKc1b+droiIn2P0Ijk5uayG/4GSk5OjvD+B10wqBq1Q4cOOHv2LNq3by+M4GI0bXiex/3799GhQwe9jsNqVEaj8FLVqMpB2q5Dd8DYxAxiqQRSIwnEksqPxEgKiUQCqVQCsVQMiUQMiZEUYom4Mk0ihkQigtRI/FeaGBKxGOK/0iRiESRSUeVfiQhSqQgSMSCRAhKRCGIxYCQFxGJAKiGIxYBETJCKAbGIIJUQJCKCWEyQiPnKbREPqZiHWMRDIuIhVqaBgxg8JKLKvyIRBwnPQQwOElJAzHMQEyf8FXEVEAnbCoiUaRwH8H+lcRzAKQBeAXAcoPhrX6EA8RyIUwAKDqTgQIoKEMeBOA58xV9pHAe+ojKdK+fAKzgQz4MrrwCv4MGVK8ArePAKHgq5AjzHQSGvtOcVBMUzBbhyHhVPOHDPePBPOQy4kYguXbroPcDeoISqHIEllbaA1KjlX0KVQiyVQFJVqEZSFaFKJOK/BC3+a7vK3+eEKjWqLlSpFJCIqwtVohSqRFWoklqEKhVV/q3cVhWqWMRBwisgpueFWpkmVhFq5baqUCv+FiqnFKrib6FySqEq/hKq+G9xSirTeI4DLxZVClWkAC8RgzgeHABezENBIvAiDryIh4IHeJEICk4EnkTgiIdCQuAkIlSIAU4kAieCsFCIvtOWWEOPYRAwoTIMAoO69Cv7fQpFGSQVRhCTBIAEYl4CXiIBT1LwEgmIl0DMicFLxOB5KTiJGDz3dxuV5yrbqJzi70t/5bYIioq626iKl7mNWlGljVrxVxtVUVmO5wgKjgPH8ajgOXDEgydOuOf9Ut1Hzc7OZstcGihZWVmwtbXV2d6gatS2bdsCAO7cuQMzM7MG9aWcTZCTk6PzyoFN1V9j+ioqKoK1tbXw2+mKQQlVeZPfzMysUcQDVPZaG8tXY/trTF/6PqBhnSmGQcCEyjAIDEqoxsbGWLp0aaNM+GtMX43tzxB9GVSvn/HyYlA1KuPlhQmVYRAwoTIMgiYr1Pj4eDg5OUEsFiM9Pb3GMkSEGTNmoEePHnBzc0NWVpZOvp49e4bx48ejR48e8PX1RX5+frUymzdvxiuvvAJnZ2c4Oztjz549OvkCKs/Nzs4OPXv2xMaNG6vlnz17Fo6OjujRo0etb0usL19du3aFTCaDs7MzRowYobOfgIAAmJubIygoqMZ8vc+JmiiZmZmUkZFBPj4+lJaWVmOZPXv2UGBgIBERxcbGCtvasnbtWpo3bx4REa1evVrYrsqmTZtqTNeWiooK6tmzJ+Xm5lJxcTH16NGDCgoKVMq4ublRamoqVVRUkJubW63nXx++bGxsqKSkROfzUXLs2DGV3+N59D2nJluj9uzZE/b29nWW2bt3L6ZMmQIAGD16NJKTk3Ua/FD1OFOnTsXevXu1D1hDlDWLlZUVTE1NMWLECBw6dEjIz8vLg0KhgEwmg1Qqxeuvv65zPOp81Se+vr61Do6uj3NqskLVhLy8PFhZWQGofETXtm1bFBQU6HUcc3NzPH78uMZyv/76K2QyGaZOnYpHjx7pHTMAdO7cGXfv3tU4vz59AZUDmr29vdG/f3/s2rVLJz/1EYc6DFqoNdWeuowk16QWHj16NLKzs5Gamgo7OzvMmzdPaz+1+aoac32dk6bHSk5OxsWLFxETE4MFCxbg5s2bOvnSNw51NCmhrl27VuislJeXqy1vZWUl/GfyPI9Hjx5pPEqnqq/27dsLxyksLESbNm2qlbewsICxsTFEIhGmTZuGc+fOaX5itcQMVE5Y7Nixo8b59ekLgPBaps6dO2Po0KG4fPmyTr70jUMterWgG4G6OlNxcXEqnanx48fr5GPNmjUqnam5c+dWK/Pnn38K25GRkRQSEqKTr4qKCurRo4dKByc/P1+ljKurq9DxcHd3pytXrjSIrydPnlBxcTERERUWFpKDgwOlp6fr5IuIKDExsdbOlL7n1GSFevDgQbKysqJmzZpR+/btaeLEiURUKc7FixcTERHHcfTOO++Qra0tubi4UGZmpk6+ysrKaOzYsdS9e3fy9vamBw8eVPMVFhZGjo6OJJPJaPjw4ZSbm6vzucXFxVHPnj2pe/fu9P333xMR0fDhw+nu3btERHT69GlycHAgW1tbWrp0qc5+1PnKysoimUxGMpmMnJycKDIyUmc/w4YNI0tLSzIxMSErKys6e/ZsvZ4Te9bPMAiaVBuVwagNJlSGQcCEyjAImFAZBgETKsMgYEJlGARMqAyDgAmVYRAwoTIMAiZUhkHAhMp4oVy6dAkTJkzAokWL6izXpIRqaWmptszcuXPh5OSEFStWNGgsb775JuLj42vM27hxI4yNjWscYF1bXkxMDDw9PeHq6oo+ffpg/vz5Go2DdXNzA8dx1dJv374NNzc3jc6lNurjGPoes1+/fli5cqXack1KqJrw008/ITU1FQsXLlRb9vkfuKYfXBeio6PRp08f7N69W6O8zZs3IyIiArGxsbhw4QLOnj0LQLPBw+7u7kJ5QyQtLQ2jRo1S+eiEzuO6GgALCwu6desWyWQymjp1Ktnb21NISAjxPE9ERIGBgSSRSKhv3760b98++vHHH8nd3Z369OkjDMdT2k+bNo369OlD169fV9l/9uxZjXZERIsWLSJ7e3saMWIEDR8+nPbu3VstxocPH5K1tTUdO3aMXnvtNbV5jx8/pnbt2tHt27d1+k5iYmJo2bJl1dJv3bpFrq6uRER05coVcnFxoaysLCIi2rBhAzk5OZFMJqOwsDAiIho1ahS5uLiQo6Mj7dq1q9oxnj+2TCaj0NBQ6t27N82YMYN2795N/fv3JwcHB2E45X/+8x9ydHQkJycn2rp1a53HrItbt27RwoUL6yzTJIVqZGREGRkZxPM8+fj4UFJSkkoZIqKrV69SUFAQKRQK4jiORo0aRadOnaJbt26RRCKh1NRUIqJq+7XZpaSkkJubG8nlcsrLyyMzM7MahRoZGUmzZs0ijuOoc+fOKgORa8rbunUrjRkzRufvpKioiHx8fKqlKwWRmpqqItLU1FRycnKix48fExEJs06Vfx8/fkx2dnbE83ydQjUyMqLr16+TQqEge3t7mj9/PhERfffdd/TBBx/QuXPnyMXFhZ4+fUoFBQVka2tLd+/e1Vqo2dnZ9Prrr5OLiwtt2bKl1nJN8tJvZ2cHe3t7iEQi9OvXD7dv365WJiEhAadPn4arqytcXFyQkZEhzOvv1asXZDKZULbqfm12p06dQkBAAJo1a4aOHTtiyJAhNcYWFRWF4OBgiMVijB49WuUSX1PetWvX4OTkJJQJCQmBvb093n77bY2+C+X6pUVFRdXy8vLyMGHCBOzYsUNYzfn48eOYMGGCsNCxcmrO119/jb59+8Lb2xt37tzBvXv36vRrZ2cHOzs7SCQS9O7dG35+fgAAmUyG27dv4+TJkwgMDETz5s3Rtm1bDB06VKfpOd26dcMvv/yCCxcuYPLkybWWa5JCrbrym0QiqbFtSUSYPn06Ll++jMuXL+PmzZvCibZo0UKlbNX92uyISG2b8f79+8jMzISXlxcAICgoCFFRUXXmmZiYqBwjOjoa06dPh42NjZA2Y8aMOv0OGTIECQkJ1dLNzc3xyiuvqG3DJiYmIjk5GWfOnEFqaiqsra0hl8vrtKn6G4jFYmFfLBbX+nvo+4qeumiSQtWEIUOGICoqCoWFhQAqJ4xpMlW6NjsvLy/s3r0b5eXluHfvHhITE6vZ7ty5E+PGjRNWT/bx8cHVq1fx8OHDWvNeffVV7NixAw8fPgQAVFRUICkpCf369QNQuUx5RkYGFixYgJkzZ9YY87Bhw3D48OFq6SYmJoiLi8OqVauE+frK81PWwI8ePUJxcTEsLCxgYmKCs2fPIjMzU+33pI5BgwYhJiYGcrkchYWFSExMhLu7u97HrQ2DWhq9Kk5OTggLC8PgwYPB8zxMTU2xfft2ne369++PV199FTKZDHZ2dvD29q5mGx0djatXr6rctiosLMSuXbtqzbtw4QLCwsLg6+uLZs2awcjICEOHDsXgwYMBABcvXsSUKVMwbdo0vPHGGzXG7O7uXquI27Rpg71792LYsGEwMzPDgAEDMHv2bHh5eUEqlWL48OFYunQpvv32Wzg7O6Nv377o06eP2u9JHW5ubggODoarqytEIhHCw8PRsWPHGptp9YLGrV5Gg/DVV19RSkoKKRQKmjZtWq3lgoKC6MaNG40YWdOCTe57wbz33nswNTVFYWEhZs6cib59+77okJokTKgMg8BgO1OMlwsmVIZBwITKMAiYUBkGARMqwyBgQmUYBEyoDIOACZVhEDChMgwCJlSGQcCEyjAI/h/WU0mQ1nMT9gAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 140x160 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "%config InlineBackend.print_figure_kwargs = {'bbox_inches':'tight'}\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from matplotlib.patches import Rectangle\n",
    "import numpy as np\n",
    "\n",
    "coefs = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2_sparseregularized/task_4/weights/weights_Binding.txt',\n",
    "                       sep='\\t')\n",
    "\n",
    "coefs.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in coefs.iterrows():\n",
    "    if row['id'] != 'WT':\n",
    "        coefs.at[index,'order']=len(row.id.split('_'))\n",
    "    else:\n",
    "        coefs.at[index,'order']=0\n",
    "coefs_1stor=coefs.copy()\n",
    "\n",
    "\n",
    "firstorderterms=coefs.loc[coefs['order']==1].copy()\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    firstorderterms.at[index,'mut_from']=row.id[0:-1]\n",
    "    firstorderterms.at[index,'wt_res']=row.id[0]\n",
    "    firstorderterms.at[index,'mut_pos']=row.id[1:-1]\n",
    "    firstorderterms.at[index,'mut_to']=row.id[-1]\n",
    "\n",
    "real_pos = [2,16,18,24,26,48,53]\n",
    "mutated_positions = []\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    if row.id !='WT' and row.mut_from not in mutated_positions:\n",
    "        mutated_positions.append(row.mut_from)\n",
    "\n",
    "coefs_hm=pd.DataFrame(columns=mutated_positions)\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    if row.id !='WT':\n",
    "        coefs_hm.at[row.mut_to,row.mut_from]=row.mean_kcal_mol\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(1.4,1.6))\n",
    "\n",
    "singles_boundary=max(abs(coefs['mean_kcal_mol']))\n",
    "\n",
    "ax=sns.heatmap(coefs_hm.fillna(np.nan), cmap='coolwarm', vmin=-1*singles_boundary, vmax=singles_boundary,cbar_kws={\"orientation\": \"horizontal\", \"pad\": 0.02, 'location':'bottom'}, xticklabels=True, yticklabels=True)\n",
    "\n",
    "ax.figure.axes[-1].set_xticks(ticks=np.arange(-1,1.5,0.5),labels=np.arange(-1,1.5,0.5),fontsize=7)\n",
    "ax.figure.axes[-1].set_xlabel(r'Inferred ${ΔΔG_{b}}$ \\ kcal mol$^-$$^1$', rotation=0, labelpad=5, fontsize=7)\n",
    "ax.spines[['right','left','top','bottom']].set_visible(True)\n",
    "ax.figure.axes[-1].spines[['right','left','top','bottom']].set_visible(True)\n",
    "plt.tick_params(axis='both', which='both', labelsize=7, labelbottom = False, bottom=False, top = True, labeltop=True)\n",
    "plt.yticks(fontsize=6)\n",
    "plt.xticks(fontsize=6, rotation=45)\n",
    "plt.xlabel('WT residue',  fontsize=7)\n",
    "ax.xaxis.set_label_position('top') \n",
    "plt.ylabel('Substitution AA', fontsize=7)\n",
    "\n",
    "for i in mutated_positions:\n",
    "    for j in mutated_positions:\n",
    "        if i[0]==j[0]:\n",
    "            col=coefs_hm.columns.get_loc(i)\n",
    "            row=coefs_hm.index.get_loc(j[0])\n",
    "            ax.add_patch(Rectangle((col, row), 1, 1, fill=True,edgecolor='black', lw=0.0001, facecolor='black'))\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_binding_1storderterms_2ndordersprsreg3statemodel_ddGs.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "05ab6f3d-daac-4cd8-86d5-fc34445d30bb",
   "metadata": {},
   "source": [
    "## Panel d"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d0c25bf0-284a-43fb-b219-2a11e4e27757",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 140x160 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%config InlineBackend.print_figure_kwargs = {'bbox_inches':'tight'}\n",
    "\n",
    "from pylab import *\n",
    "import matplotlib as mpl\n",
    "import itertools\n",
    "mpl.rc('font',family='Arial')\n",
    "\n",
    "def flip(items, ncol):\n",
    "    return itertools.chain(*[items[i::ncol] for i in range(ncol)])\n",
    "\n",
    "dist_core_lig = [11.916178703308105, 9.267768859863281, 7.30642032623291, 12.082030296325684, 11.4977388381958, 8.362347602844238, 8.694024085998535] # minimum side chain heavy atom distances to ligand \n",
    "#dist_core_lig = [11.916177749633789, 7.154636859893799, 7.30642032623291, 12.082030296325684, 11.4977388381958, 5.728403568267822, 7.581732749938965] # minimum all atom distances to ligand\n",
    "# legend label ordered by distance\n",
    "pos_df = pd.DataFrame({'dist':dist_core_lig}, index=coefs_hm.columns)\n",
    "pos_dist_asc = pos_df.sort_values(by='dist').index.to_list()\n",
    "\n",
    "y = coefs_hm.T.to_numpy().flatten()\n",
    "listoflists = [np.repeat(dist_core_lig[i], 5) for i in range(0,7)]\n",
    "x = [item for list in listoflists for item in list]\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(1.4,1.6))\n",
    "ax.axhspan(-1, coefs_hm.mean().mean(), facecolor='gray', alpha=0.1)\n",
    "sns.scatterplot(x=x,y=y, hue=np.repeat(coefs_hm.columns, 5), s=0, palette='Dark2')\n",
    "sns.scatterplot(x=dist_core_lig,y=coefs_hm.mean(), palette='Dark2',s=50,marker='o', alpha=0.3, hue=list(pos_df.index))\n",
    "#plt.axhline(coefs_hm.mean().mean(), linestyle='--', color='red', linewidth=1)\n",
    "#plt.axvline(5, linestyle='--', color='black', linewidth=1)\n",
    "h, l = ax.get_legend_handles_labels()\n",
    "plt.legend(flip(h[0:7],4),flip(l[0:7],4),title = 'WT residue', fontsize=6, title_fontsize=7, frameon=False, ncol=(len(pos_df.index)+1)/2,bbox_to_anchor=(1.1,1.37), columnspacing=0, handletextpad=0)\n",
    "plt.xlabel('Distance to ligand\\nÅ', fontsize=7)\n",
    "plt.ylabel(r'|${ΔΔG_{b}}$|'+ '\\n'+ r'kcal mol$^-$$^1$', fontsize=7)\n",
    "plt.xticks(fontsize=6)\n",
    "plt.yticks(fontsize=6)\n",
    "ax.set_xticks(np.arange(5,15,2))\n",
    "#ax.set_yticks(np.arange(-1.5,0.5,0.5))\n",
    "ax.set_ylim(-0.3,1.3)\n",
    "ax.spines[['top','right']].set_visible(False)\n",
    "#ax.get_legend().remove()\n",
    "\n",
    "# TODO jitter label\n",
    "def label_point(x, y, val, ax, colors):\n",
    "    a = pd.concat({'x': x, 'y': y, 'val': val}, axis=1)\n",
    "    col = a.iloc[0]['x']\n",
    "    counter=0\n",
    "    for i, point in a.iterrows():\n",
    "        if point['x'] != col:\n",
    "            col=point['x']\n",
    "            counter+=1\n",
    "        ax.text(point['x'], point['y'], str(point['val']), fontsize=6, color=colors[counter], ha='center')\n",
    "\n",
    "colors=list(sns.color_palette('Dark2',len(pos_df)))\n",
    "\n",
    "\n",
    "label_point(pd.Series(x).loc[pd.Series(y).notnull()], pd.Series(y).loc[pd.Series(y).notnull()], pd.Series(np.tile(coefs_hm.index, 7)).loc[pd.Series(y).notnull()], plt.gca(), colors)\n",
    "\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_binding_ddGs_2ndordersprsreg3statemodel_vs_distance_scatter.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9f4e3316-4dc4-4823-9bbb-16ec1393a6c7",
   "metadata": {},
   "source": [
    "## Panel e"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "1d02f74f-f3e9-47f6-a80f-7f887dc31d40",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2_sparseregularized/task_4/weights/weights_Binding.txt',\n",
    "                       sep='\\t')\n",
    "stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in stability_weights_3state_2ndorder_sprsreg.iterrows():\n",
    "    if row['id'] != 'WT':\n",
    "        stability_weights_3state_2ndorder_sprsreg.at[index,'order']=len(row.id.split('_'))\n",
    "    else:\n",
    "        stability_weights_3state_2ndorder_sprsreg.at[index,'order']=0\n",
    "\n",
    "stability_weights_3state_1storder=pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order1/task_1/weights/weights_Binding.txt',\n",
    "                   sep='\\t')\n",
    "\n",
    "for index, row in stability_weights_3state_1storder.iterrows():\n",
    "    if row.id != 'WT':\n",
    "        stability_weights_3state_1storder.at[index,'order']=len(row.id_ref.split('_'))\n",
    "    else:\n",
    "        stability_weights_3state_1storder.at[index,'order']=0\n",
    "\n",
    "stability_weights_3state_1storder.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "eea82eda-9311-468a-98bb-6ad3d146195c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_3state_1storder = stability_weights_3state_1storder.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==1].merge(stability_weights_3state_1storder.loc[stability_weights_3state_1storder['order']==1], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔG_{b}}$ from sparse regularized'+'\\n2nd order 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔG_{b}}$ from 1st order 3-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storderterms_1storder3statemodel_vs_1storder2statemodel_FoldingddGs_TrueWT_recalibrated.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4025a9c3-563b-4477-8fdc-40750fd6ad7d",
   "metadata": {},
   "source": [
    "## Panel f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "a1fe2eb9-c01d-49f8-8646-6ee76b50f556",
   "metadata": {},
   "outputs": [],
   "source": [
    "stability_weights_3state_2ndorder=pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2/task_1/weights/weights_Binding.txt',\n",
    "                   sep='\\t')\n",
    "\n",
    "for index, row in stability_weights_3state_2ndorder.iterrows():\n",
    "    if row.id != 'WT':\n",
    "        stability_weights_3state_2ndorder.at[index,'order']=len(row.id_ref.split('_'))\n",
    "    else:\n",
    "        stability_weights_3state_2ndorder.at[index,'order']=0\n",
    "\n",
    "stability_weights_3state_2ndorder.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "abf50caa-2ff9-4e26-8975-28fa025dd42f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_3state_2ndorder = stability_weights_3state_2ndorder.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==1].merge(stability_weights_3state_2ndorder.loc[stability_weights_3state_2ndorder['order']==1], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔG_{b}}$ from sparse regularized'+'\\n2nd order 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔG_{b}}$ from 2nd order 3-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storderterms_1storder3statemodel_vs_1storder2statemodel_FoldingddGs_TrueWT_recalibrated.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "16b13422-aaec-426d-a27c-a2b156e4220b",
   "metadata": {},
   "source": [
    "## Panel g"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "c0555479-1265-4c6e-ac0b-be3db455734b",
   "metadata": {},
   "outputs": [],
   "source": [
    "coefs = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2_sparseregularized/task_4/weights/weights_Binding.txt',\n",
    "                       sep='\\t')\n",
    "coefs.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in coefs.iterrows():\n",
    "    coefs.at[index,'order']=len(row.id.split('_'))\n",
    "singles_boundary=max(abs(coefs['mean_kcal_mol']))\n",
    "firstorderterms=coefs.loc[coefs['order']==1].copy()\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    firstorderterms.at[index,'mut_from']=row.id[0:-1]\n",
    "    firstorderterms.at[index,'wt_res']=row.id[0]\n",
    "    firstorderterms.at[index,'mut_pos']=row.id[1:-1]\n",
    "    firstorderterms.at[index,'mut_to']=row.id[-1]\n",
    "\n",
    "real_pos = [2,16,18,24,26,48,53]\n",
    "mutated_positions = []\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    if row.id !='WT' and row.mut_from not in mutated_positions:\n",
    "        mutated_positions.append(row.mut_from)\n",
    "\n",
    "secondorderterms=coefs.loc[coefs['order']==2].copy()\n",
    "\n",
    "wt ='FLFFIIV'\n",
    "mut_pos_name = [2,16,18,24,26,48,53]\n",
    "\n",
    "\n",
    "secondorderterms['mutation1']='a'\n",
    "secondorderterms['mutation2']='a'\n",
    "for index,row in secondorderterms.iterrows():\n",
    "    secondorderterms.at[index,'mutation1']=row.id.split('_')[0]\n",
    "    secondorderterms.at[index,'mutation2']=row.id.split('_')[1]\n",
    "\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    secondorderterms.at[index,'pos1']=int(row.mutation1[1:-1])\n",
    "    secondorderterms.at[index,'pos2']=int(row.mutation2[1:-1])\n",
    "secondorderterms=secondorderterms.sort_values(['pos1','pos2'])\n",
    "\n",
    "single_mutations=[]\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    if row.mutation1 not in single_mutations:\n",
    "        single_mutations.append(row.mutation1)\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    if row.mutation2 not in single_mutations:\n",
    "        single_mutations.append(row.mutation2)\n",
    "\n",
    "couplings_boundary=max(abs(secondorderterms['mean_kcal_mol']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "a48e96c4-5e19-483c-beb9-63ae0b3e2b98",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 300x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.patches import Rectangle\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import math\n",
    "#from matplotlib.colors import TwoSlopeNorm\n",
    "\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "\n",
    "pairwise_secondorderterms=pd.DataFrame(columns=single_mutations)\n",
    "\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    pairwise_secondorderterms.at[row.mutation1,row.mutation2]=row.mean_kcal_mol\n",
    "\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    pairwise_secondorderterms.at[row.mutation2,row.mutation1]=row.mean_kcal_mol\n",
    "\n",
    "pairwise_secondorderterms=pairwise_secondorderterms.reindex(single_mutations)\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(3,3))\n",
    "for mutation_nr, mutation in enumerate(mutated_positions[1:]):\n",
    "    plt.text(-4.2, 2.5+4*mutation_nr, mutation, fontsize=8)\n",
    "for mutation_nr, mutation in enumerate(mutated_positions[:-1]):\n",
    "    plt.text(1+4*mutation_nr, 27.2, mutation, fontsize=8)\n",
    "\n",
    "\n",
    "# Create a custom mask\n",
    "mask = np.zeros_like(pairwise_secondorderterms.iloc[4:,:-4], dtype=bool)\n",
    "for i in range(len(pairwise_secondorderterms.iloc[4:,:-4].index)):\n",
    "    for j in range(i+1, len(pairwise_secondorderterms.iloc[4:,:-4].index)):  # Only mask rows and columns in the upper diagonal\n",
    "        if (i // 4) < (j // 4):  # Mask entire quadrants in the upper diagonal\n",
    "            mask[i, j] = True\n",
    "\n",
    "\n",
    "couplings_boundary=max(abs(secondorderterms['mean_kcal_mol']))\n",
    "cbar_ax=fig.add_axes([0.3,0.85,0.3,0.02])\n",
    "hmap=sns.heatmap(pairwise_secondorderterms.iloc[4:,:-4].fillna(np.nan), cmap='coolwarm', center=0, xticklabels=True,yticklabels=True, mask=pairwise_secondorderterms.iloc[4:,:-4].fillna(np.nan).isnull()+mask,vmin=-1*couplings_boundary, vmax=couplings_boundary,\n",
    "              cbar_kws={\"orientation\": \"horizontal\", \"pad\": 0.02, 'ticks':list(np.arange(-1,1.5,0.5))}, ax=ax, cbar_ax=cbar_ax,  fmt= '', annot_kws={'size':4})\n",
    "\n",
    "\n",
    "cbar_ax.set_xlabel(r'Inferred ${ΔΔΔG_{b}}$ \\ kcal mol$^-$$^1$', rotation=0, fontsize=8)\n",
    "cbar_ax.xaxis.set_label_position('top')\n",
    "for spine in cbar_ax.spines.values():\n",
    "    spine.set(visible=True,lw=1,edgecolor='black')\n",
    "plt.xticks(fontsize=6)\n",
    "plt.yticks(fontsize=6)\n",
    "    \n",
    "labels_x = [item.get_text() for item in ax.get_xticklabels()]\n",
    "for label_nr, label in enumerate(labels_x):\n",
    "    labels_x[label_nr]=label[-1]\n",
    "ax.set_xticklabels(labels_x, rotation=0,fontsize=6)\n",
    "labels_y = [item.get_text() for item in ax.get_yticklabels()]\n",
    "for label_nr, label in enumerate(labels_y):\n",
    "    labels_y[label_nr]=label[-1]\n",
    "ax.set_yticklabels(labels_y, rotation=0,fontsize=6)\n",
    "\n",
    "for i in range(0, len(pairwise_secondorderterms.iloc[4:,:-4].index), 4):\n",
    "    for j in range(0, len(pairwise_secondorderterms.iloc[4:,:-4].index), 4):\n",
    "        if not np.any(mask[i:i+4, j:j+4]):\n",
    "            rect = Rectangle((j, i), 4, 4, linewidth=1, edgecolor='black', facecolor='none')\n",
    "            ax.add_patch(rect)\n",
    "\n",
    "ax.axhline(y=len(pairwise_secondorderterms.iloc[4:,:-4].index), color='k',linewidth=2)\n",
    "ax.axhline(y=0,xmin=0,xmax=0.16, color='k',linewidth=2)\n",
    "ax.axvline(x=0, color='k',linewidth=2)\n",
    "ax.axvline(x=len(pairwise_secondorderterms.iloc[4:,:-4].index),ymin=0,ymax=0.16, color='k',linewidth=2)\n",
    "    \n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_SH3_2or_pairwisecoup_sparsereg.svg', dpi=300, transparent=True, bbox_inches = \"tight\")\n",
    "            "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ecf83629-fab7-40a5-9a6f-ffcd4b9b1bae",
   "metadata": {},
   "source": [
    "## Panel h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "25d91371-9de3-4a58-9f53-e04b1d1f9393",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_3state_2ndorder = stability_weights_3state_2ndorder.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==2].merge(stability_weights_3state_2ndorder.loc[stability_weights_3state_2ndorder['order']==2], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔΔG_{b}}$ from sparse regularized'+'\\n2nd order 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔΔG_{b}}$ from 2nd order 3-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59a356d5-7e7c-4bed-b692-26f878d17ce9",
   "metadata": {},
   "source": [
    "## Panel i"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "09e96895-7373-4e45-b513-df872f3fd072",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 140x160 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "\n",
    "sparse_model_performances_stability = pd.DataFrame()\n",
    "\n",
    "y = predicted_abundance_scaled.loc[predicted_abundance_scaled['WT']!=True][['stability_fitness']].to_numpy()\n",
    "x = predicted_abundance_scaled.loc[predicted_abundance_scaled['WT']!=True][['predicted_stability_fitness']].to_numpy()\n",
    "model = LinearRegression()\n",
    "model.fit(x,y)\n",
    "y_pred = model.predict(x)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(1.4,1.6))\n",
    "hbin=ax.hexbin(x=predicted_abundance_scaled['predicted_stability_fitness'].loc[predicted_abundance_scaled['WT']!=True],y=predicted_abundance_scaled['stability_fitness'].loc[predicted_abundance_scaled['WT']!=True],bins='log',gridsize=75,rasterized=True,linewidths=0,vmax=100)\n",
    "ax.axline((0.1, 0.1), slope=1, linestyle = \"dashed\", color = 'black',lw=0.5)\n",
    "\n",
    "\n",
    "plt.axvline(0, ls='--', color='black', lw=0.5)\n",
    "plt.axhline(0, ls='--', color='black', lw=0.5)\n",
    "plt.title('\\nR$^2$ = '+str(round(r2_score(y,y_pred),2)),fontsize=7, color='r', loc='right')\n",
    "sparse_model_performances_stability.at['2','All']=round(r2_score(y,y_pred),2)\n",
    "plt.xlabel('Abundance fitness\\n(predicted)', fontsize=7)\n",
    "plt.xticks(fontsize=6)\n",
    "plt.ylabel('Abundance fitness\\n(observed)', fontsize=7)\n",
    "plt.yticks(fontsize=6)\n",
    "ax.set_xticks(np.arange(-1.5,0.5,0.5))\n",
    "ax.set_yticks(np.arange(-1.5,0.5,0.5))\n",
    "ax.set_ylim(-1.3,0.2)\n",
    "ax.set_xlim(-1.05,0.2)\n",
    "ax.spines[['top','right']].set_visible(False)\n",
    "ax.set_rasterization_zorder(0)\n",
    "fmt = '%1.f'\n",
    "cax=fig.add_axes([0.15,0.82,0.3,0.05])\n",
    "cb=fig.colorbar(hbin, ax=ax, shrink=0.7,format=fmt,location='top',cax=cax)\n",
    "cb.outline.set_visible(False)\n",
    "cb.ax.tick_params(axis='both', which='both',direction='in', color='white',width=0.25)\n",
    "cb.ax.xaxis.set_tick_params(pad=0)\n",
    "#cb.ax.set_title('Count', fontsize=7)\n",
    "#cb.set_label('count', rotation=0, fontsize=7, labelpad=2)\n",
    "#cb.ax.xaxis.set_label_position('right')\n",
    "cb.set_ticks([1,10,100])\n",
    "cb.set_ticklabels([1,10,100], fontsize=6)\n",
    "ax1 = cb.ax\n",
    "ax1.text(140,0.1,'count',rotation=0, fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_2ndordersprsreg_abundance_performance_bindingPCA.svg', dpi=300, transparent=True, bbox_inches='tight')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3f4cd7cd-8e01-4ce4-b834-cfabc82263fa",
   "metadata": {},
   "source": [
    "## Panel j"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "f2f05fa1-d463-4f78-ad69-c652699ef0f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 140x160 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "%config InlineBackend.print_figure_kwargs = {'bbox_inches':'tight'}\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from matplotlib.patches import Rectangle\n",
    "import numpy as np\n",
    "\n",
    "coefs = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2_sparseregularized/task_4/weights/weights_Folding.txt',\n",
    "                       sep='\\t')\n",
    "\n",
    "coefs.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in coefs.iterrows():\n",
    "    if row['id'] != 'WT':\n",
    "        coefs.at[index,'order']=len(row.id.split('_'))\n",
    "    else:\n",
    "        coefs.at[index,'order']=0\n",
    "coefs_1stor=coefs.copy()\n",
    "\n",
    "\n",
    "firstorderterms=coefs.loc[coefs['order']==1].copy()\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    firstorderterms.at[index,'mut_from']=row.id[0:-1]\n",
    "    firstorderterms.at[index,'wt_res']=row.id[0]\n",
    "    firstorderterms.at[index,'mut_pos']=row.id[1:-1]\n",
    "    firstorderterms.at[index,'mut_to']=row.id[-1]\n",
    "\n",
    "real_pos = [2,16,18,24,26,48,53]\n",
    "mutated_positions = []\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    if row.id !='WT' and row.mut_from not in mutated_positions:\n",
    "        mutated_positions.append(row.mut_from)\n",
    "\n",
    "coefs_hm=pd.DataFrame(columns=mutated_positions)\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    if row.id !='WT':\n",
    "        coefs_hm.at[row.mut_to,row.mut_from]=row.mean_kcal_mol\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(1.4,1.6))\n",
    "\n",
    "singles_boundary=max(abs(coefs['mean_kcal_mol']))\n",
    "\n",
    "ax=sns.heatmap(coefs_hm.fillna(np.nan), cmap='coolwarm', vmin=-1*singles_boundary, vmax=singles_boundary,cbar_kws={\"orientation\": \"horizontal\", \"pad\": 0.02, 'location':'bottom'}, xticklabels=True, yticklabels=True)\n",
    "\n",
    "ax.figure.axes[-1].set_xticks(ticks=np.arange(-1,1.5,0.5),labels=np.arange(-1,1.5,0.5),fontsize=7)\n",
    "ax.figure.axes[-1].set_xlabel(r'Inferred ${ΔΔG_{f}}$ \\ kcal mol$^-$$^1$', rotation=0, labelpad=5, fontsize=7)\n",
    "ax.spines[['right','left','top','bottom']].set_visible(True)\n",
    "ax.figure.axes[-1].spines[['right','left','top','bottom']].set_visible(True)\n",
    "plt.tick_params(axis='both', which='both', labelsize=7, labelbottom = False, bottom=False, top = True, labeltop=True)\n",
    "plt.yticks(fontsize=6)\n",
    "plt.xticks(fontsize=6, rotation=45)\n",
    "plt.xlabel('WT residue',  fontsize=7)\n",
    "ax.xaxis.set_label_position('top') \n",
    "plt.ylabel('Substitution AA', fontsize=7)\n",
    "\n",
    "for i in mutated_positions:\n",
    "    for j in mutated_positions:\n",
    "        if i[0]==j[0]:\n",
    "            col=coefs_hm.columns.get_loc(i)\n",
    "            row=coefs_hm.index.get_loc(j[0])\n",
    "            ax.add_patch(Rectangle((col, row), 1, 1, fill=True,edgecolor='black', lw=0.0001, facecolor='black'))\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_folding_1storderterms_2ndordersprsreg3statemodel_ddGs.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90e15f77-aaa9-4dbc-a71c-8118fb7a186d",
   "metadata": {},
   "source": [
    "## Panel k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "88c71ec3-87bf-465c-9152-9c1dac36f257",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2_sparseregularized/task_4/weights/weights_Folding.txt',\n",
    "                       sep='\\t')\n",
    "stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in stability_weights_3state_2ndorder_sprsreg.iterrows():\n",
    "    if row['id'] != 'WT':\n",
    "        stability_weights_3state_2ndorder_sprsreg.at[index,'order']=len(row.id.split('_'))\n",
    "    else:\n",
    "        stability_weights_3state_2ndorder_sprsreg.at[index,'order']=0\n",
    "\n",
    "stability_weights_2state_2ndorder_sprsreg=pd.read_csv(path+'Fig3/FYN-SH3/FYN_mochi_order2_sparse_regularized/task_4/weights/weights_Folding.txt',\n",
    "                   sep='\\t')\n",
    "\n",
    "for index, row in stability_weights_2state_2ndorder_sprsreg.iterrows():\n",
    "    if row.id != 'WT':\n",
    "        stability_weights_2state_2ndorder_sprsreg.at[index,'order']=len(row.id_ref.split('_'))\n",
    "    else:\n",
    "        stability_weights_2state_2ndorder_sprsreg.at[index,'order']=0\n",
    "\n",
    "stability_weights_2state_2ndorder_sprsreg.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "c63a9fc4-2861-428a-8f79-e38fad38eb5a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_2state_2ndorder_sprsreg = stability_weights_2state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==1].merge(stability_weights_2state_2ndorder_sprsreg.loc[stability_weights_2state_2ndorder_sprsreg['order']==1], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔG_{f}}$ from 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔG_{f}}$ from 2-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storderterms_1storder3statemodel_vs_1storder2statemodel_FoldingddGs_TrueWT_recalibrated.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4441dbd-5c77-48a8-9c24-d020f51113b3",
   "metadata": {},
   "source": [
    "## Panel l"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "34e610e4-9796-4474-a2ef-48f44c01c783",
   "metadata": {},
   "outputs": [],
   "source": [
    "stability_weights_3state_2ndorder = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2/task_1/weights/weights_Folding.txt',\n",
    "                       sep='\\t')\n",
    "stability_weights_3state_2ndorder.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in stability_weights_3state_2ndorder.iterrows():\n",
    "    if row['id'] != 'WT':\n",
    "        stability_weights_3state_2ndorder.at[index,'order']=len(row.id.split('_'))\n",
    "    else:\n",
    "        stability_weights_3state_2ndorder.at[index,'order']=0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "0a92d43f-a284-4f01-a677-923ad10783d3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_3state_2ndorder = stability_weights_3state_2ndorder.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==1].merge(stability_weights_3state_2ndorder.loc[stability_weights_3state_2ndorder['order']==1], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔG_{f}}$ from sparse regularized'+'\\n2nd order 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔG_{f}}$ from 2nd order 3-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storderterms_1storder3statemodel_vs_1storder2statemodel_FoldingddGs_TrueWT_recalibrated.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "15baecd1-f730-4a2e-998c-dbc899b0ff12",
   "metadata": {},
   "source": [
    "## Panel m"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "7eb56fb3-6d51-49b8-b7f8-4256afbb79e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "coefs = pd.read_csv(path+'Fig7/'+'FYN_coreDTS_bPCA_order2_sparseregularized/task_4/weights/weights_Folding.txt',\n",
    "                       sep='\\t')\n",
    "coefs.rename(columns={'mean_kcal/mol':'mean_kcal_mol', 'std_kcal/mol':'std_kcal_mol', 'ci95_kcal/mol':'ci95_kcal_mol'}, inplace=True)\n",
    "for index, row in coefs.iterrows():\n",
    "    coefs.at[index,'order']=len(row.id.split('_'))\n",
    "singles_boundary=max(abs(coefs['mean_kcal_mol']))\n",
    "firstorderterms=coefs.loc[coefs['order']==1].copy()\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    firstorderterms.at[index,'mut_from']=row.id[0:-1]\n",
    "    firstorderterms.at[index,'wt_res']=row.id[0]\n",
    "    firstorderterms.at[index,'mut_pos']=row.id[1:-1]\n",
    "    firstorderterms.at[index,'mut_to']=row.id[-1]\n",
    "\n",
    "real_pos = [2,16,18,24,26,48,53]\n",
    "mutated_positions = []\n",
    "\n",
    "for index, row in firstorderterms.iterrows():\n",
    "    if row.id !='WT' and row.mut_from not in mutated_positions:\n",
    "        mutated_positions.append(row.mut_from)\n",
    "\n",
    "secondorderterms=coefs.loc[coefs['order']==2].copy()\n",
    "\n",
    "wt ='FLFFIIV'\n",
    "mut_pos_name = [2,16,18,24,26,48,53]\n",
    "\n",
    "\n",
    "secondorderterms['mutation1']='a'\n",
    "secondorderterms['mutation2']='a'\n",
    "for index,row in secondorderterms.iterrows():\n",
    "    secondorderterms.at[index,'mutation1']=row.id.split('_')[0]\n",
    "    secondorderterms.at[index,'mutation2']=row.id.split('_')[1]\n",
    "\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    secondorderterms.at[index,'pos1']=int(row.mutation1[1:-1])\n",
    "    secondorderterms.at[index,'pos2']=int(row.mutation2[1:-1])\n",
    "secondorderterms=secondorderterms.sort_values(['pos1','pos2'])\n",
    "\n",
    "single_mutations=[]\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    if row.mutation1 not in single_mutations:\n",
    "        single_mutations.append(row.mutation1)\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    if row.mutation2 not in single_mutations:\n",
    "        single_mutations.append(row.mutation2)\n",
    "\n",
    "couplings_boundary=max(abs(secondorderterms['mean_kcal_mol']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7e4fa1b3-7578-4427-bf4a-62bc35eb9e0c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 300x300 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib.patches import Rectangle\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import math\n",
    "#from matplotlib.colors import TwoSlopeNorm\n",
    "\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "\n",
    "pairwise_secondorderterms=pd.DataFrame(columns=single_mutations)\n",
    "\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    pairwise_secondorderterms.at[row.mutation1,row.mutation2]=row.mean_kcal_mol\n",
    "\n",
    "for index, row in secondorderterms.iterrows():\n",
    "    pairwise_secondorderterms.at[row.mutation2,row.mutation1]=row.mean_kcal_mol\n",
    "\n",
    "pairwise_secondorderterms=pairwise_secondorderterms.reindex(single_mutations)\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(3,3))\n",
    "for mutation_nr, mutation in enumerate(mutated_positions[1:]):\n",
    "    plt.text(-4.2, 2.5+4*mutation_nr, mutation, fontsize=8)\n",
    "for mutation_nr, mutation in enumerate(mutated_positions[:-1]):\n",
    "    plt.text(1+4*mutation_nr, 27.2, mutation, fontsize=8)\n",
    "\n",
    "\n",
    "# Create a custom mask\n",
    "mask = np.zeros_like(pairwise_secondorderterms.iloc[4:,:-4], dtype=bool)\n",
    "for i in range(len(pairwise_secondorderterms.iloc[4:,:-4].index)):\n",
    "    for j in range(i+1, len(pairwise_secondorderterms.iloc[4:,:-4].index)):  # Only mask rows and columns in the upper diagonal\n",
    "        if (i // 4) < (j // 4):  # Mask entire quadrants in the upper diagonal\n",
    "            mask[i, j] = True\n",
    "\n",
    "\n",
    "couplings_boundary=max(abs(secondorderterms['mean_kcal_mol']))\n",
    "cbar_ax=fig.add_axes([0.3,0.85,0.3,0.02])\n",
    "hmap=sns.heatmap(pairwise_secondorderterms.iloc[4:,:-4].fillna(np.nan), cmap='coolwarm', center=0, xticklabels=True,yticklabels=True, mask=pairwise_secondorderterms.iloc[4:,:-4].fillna(np.nan).isnull()+mask,vmin=-1*couplings_boundary, vmax=couplings_boundary,\n",
    "              cbar_kws={\"orientation\": \"horizontal\", \"pad\": 0.02, 'ticks':list(np.arange(-1,1.5,0.5))}, ax=ax, cbar_ax=cbar_ax,  fmt= '', annot_kws={'size':4})\n",
    "\n",
    "\n",
    "cbar_ax.set_xlabel(r'Inferred ${ΔΔΔG_{f}}$ \\ kcal mol$^-$$^1$', rotation=0, fontsize=8)\n",
    "cbar_ax.xaxis.set_label_position('top')\n",
    "for spine in cbar_ax.spines.values():\n",
    "    spine.set(visible=True,lw=1,edgecolor='black')\n",
    "plt.xticks(fontsize=6)\n",
    "plt.yticks(fontsize=6)\n",
    "    \n",
    "labels_x = [item.get_text() for item in ax.get_xticklabels()]\n",
    "for label_nr, label in enumerate(labels_x):\n",
    "    labels_x[label_nr]=label[-1]\n",
    "ax.set_xticklabels(labels_x, rotation=0,fontsize=6)\n",
    "labels_y = [item.get_text() for item in ax.get_yticklabels()]\n",
    "for label_nr, label in enumerate(labels_y):\n",
    "    labels_y[label_nr]=label[-1]\n",
    "ax.set_yticklabels(labels_y, rotation=0,fontsize=6)\n",
    "\n",
    "for i in range(0, len(pairwise_secondorderterms.iloc[4:,:-4].index), 4):\n",
    "    for j in range(0, len(pairwise_secondorderterms.iloc[4:,:-4].index), 4):\n",
    "        if not np.any(mask[i:i+4, j:j+4]):\n",
    "            rect = Rectangle((j, i), 4, 4, linewidth=1, edgecolor='black', facecolor='none')\n",
    "            ax.add_patch(rect)\n",
    "\n",
    "ax.axhline(y=len(pairwise_secondorderterms.iloc[4:,:-4].index), color='k',linewidth=2)\n",
    "ax.axhline(y=0,xmin=0,xmax=0.16, color='k',linewidth=2)\n",
    "ax.axvline(x=0, color='k',linewidth=2)\n",
    "ax.axvline(x=len(pairwise_secondorderterms.iloc[4:,:-4].index),ymin=0,ymax=0.16, color='k',linewidth=2)\n",
    "    \n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_SH3_2or_pairwisecoup_sparsereg.svg', dpi=300, transparent=True, bbox_inches = \"tight\")\n",
    "            "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9bcb3299-3edd-4f68-8229-27761e11b7df",
   "metadata": {},
   "source": [
    "## Panel n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "5eb03340-b644-4ef0-a345-11e42bd6d74e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_2state_2ndorder_sprsreg = stability_weights_2state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==2].merge(stability_weights_2state_2ndorder_sprsreg.loc[stability_weights_2state_2ndorder_sprsreg['order']==2], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔG_{f}}$ from 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔG_{f}}$ from 2-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storderterms_1storder3statemodel_vs_1storder2statemodel_FoldingddGs_TrueWT_recalibrated.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cffaf1c8-b013-4742-b1c9-604d439e0935",
   "metadata": {},
   "source": [
    "## Panel o"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "59c28ecc-dcc2-458b-895f-7a1a3b01cf10",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 150x150 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "stability_weights_3state_2ndorder_sprsreg = stability_weights_3state_2ndorder_sprsreg.rename(columns={'mean_kcal_mol':'mean_kcal_mol_3state'})\n",
    "stability_weights_3state_2ndorder = stability_weights_3state_2ndorder.rename(columns={'mean_kcal_mol':'mean_kcal_mol_2state'})\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels = stability_weights_3state_2ndorder_sprsreg.loc[stability_weights_3state_2ndorder_sprsreg['order']==2].merge(stability_weights_3state_2ndorder.loc[stability_weights_3state_2ndorder['order']==2], left_on='id', right_on='id')\n",
    "\n",
    "singles_folding_ddGs_2statevs3statemodels=singles_folding_ddGs_2statevs3statemodels.sort_values('mean_kcal_mol_3state')\n",
    "\n",
    "x = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_3state']].to_numpy()\n",
    "y = singles_folding_ddGs_2statevs3statemodels[['mean_kcal_mol_2state']].to_numpy()\n",
    "\n",
    "model = LinearRegression()\n",
    "model.fit(x, y)\n",
    "\n",
    "y_pred = model.predict(x)\n",
    "r2 = r2_score(y,y_pred)\n",
    "r,p = pearsonr(y.ravel(),y_pred.ravel())\n",
    "\n",
    "fig, ax=plt.subplots(figsize=(1.5,1.5))\n",
    "plt.plot(x, y_pred, color='Gray', linewidth=1, linestyle='--')\n",
    "singles_folding_ddGs_2statevs3statemodels.plot.scatter('mean_kcal_mol_3state','mean_kcal_mol_2state', ax=ax, color='k', s=5)\n",
    "ax.set_xlabel(r'${ΔΔG_{f}}$ from sparse regularized'+'\\n2nd order 3-state model', fontsize=7)\n",
    "ax.set_ylabel(r'${ΔΔG_{f}}$ from 2nd order 3-state model', fontsize=7)\n",
    "ax.spines[['top', 'right']].set_visible(False)\n",
    "ax.tick_params(labelsize=6)\n",
    "plt.title('R = '+str(round(r,2))+ '\\np-val = '+'{:0.2e}'.format(p), fontsize=7)\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_Mochi_1storderterms_1storder3statemodel_vs_1storder2statemodel_FoldingddGs_TrueWT_recalibrated.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "79652053-c6ad-4dbd-b318-4ab8f52ab26f",
   "metadata": {},
   "source": [
    "## Panel p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "ab1c2bb4-7d4f-4e7c-8ca5-6e54ddd6c177",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "predicted_phenotypes = pd.read_csv(path+'FigS14'+'/FYN_coreDTS_bPCA_norm_order2_sparseregularized/task_4/predictions/predicted_phenotypes_all.txt',\n",
    "                                  sep='\\t', lineterminator='\\n', low_memory=False)\n",
    "\n",
    "predicted_abundance = predicted_phenotypes.loc[predicted_phenotypes['Abundance']==1].copy()\n",
    "predicted_binding = predicted_phenotypes.loc[predicted_phenotypes['Binding']==1].copy()\n",
    "\n",
    "linear_weights_abundance = pd.read_csv(path+'FigS14'+'/FYN_coreDTS_bPCA_norm_order2_sparseregularized/task_4/weights/'+'linears_weights_Abundance.txt',\n",
    "                                      sep='\\t')\n",
    "linear_weights_binding = pd.read_csv(path+'FigS14'+'/FYN_coreDTS_bPCA_norm_order2_sparseregularized/task_4/weights/'+'linears_weights_Binding.txt',\n",
    "                                    sep='\\t')\n",
    "weights_binding = pd.read_csv(path+'FigS14'+'/FYN_coreDTS_bPCA_norm_order2_sparseregularized/task_4/weights/'+'weights_Binding.txt',\n",
    "                             sep='\\t')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "1a6dbc9a-fdf2-4750-a052-476747fe7d58",
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import statistics\n",
    "RT=0.001987*(273+30)\n",
    "\n",
    "for i in range(1,11,1):\n",
    "    fold = i\n",
    "    \n",
    "    folding_fitness = predicted_abundance[['fitness']]\n",
    "    folding_linear_kernel = linear_weights_abundance.loc[linear_weights_abundance['fold']==fold].iloc[0]['kernel']\n",
    "    folding_linear_bias = linear_weights_abundance.loc[linear_weights_abundance['fold']==fold].iloc[0]['bias']\n",
    "    binding_linear_kernel = linear_weights_binding.loc[linear_weights_binding['fold']==fold].iloc[0]['kernel']\n",
    "    binding_linear_bias = linear_weights_binding.loc[linear_weights_binding['fold']==fold].iloc[0]['bias']\n",
    "    b_dg_wt = weights_binding.loc[weights_binding['id']=='WT'].iloc[0]['fold_'+str(fold)]*RT\n",
    "    \n",
    "    Cb = math.exp(b_dg_wt/RT)\n",
    "    predicted_abundance['binding_fitness_from_abundance_fitness_fold'+str(i)]=binding_linear_kernel/(Cb*folding_linear_kernel/(folding_fitness-folding_linear_bias)+1)+binding_linear_bias\n",
    "for index, row in predicted_abundance.iterrows():\n",
    "    binding_fitness_from_abundance_fitness = []\n",
    "    for i in range(1,11,1):\n",
    "        binding_fitness_from_abundance_fitness.append(row['binding_fitness_from_abundance_fitness_fold'+str(i)])\n",
    "    predicted_abundance.at[index,'binding_fitness_from_abundance_fitness']=statistics.mean(binding_fitness_from_abundance_fitness)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "4dbeb1db-addb-44fe-902b-06aac798a47a",
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_abundance = predicted_abundance.rename(columns={'predicted_fitness': 'predicted_stability_fitness'})\n",
    "predicted_binding = predicted_binding.rename(columns={'predicted_fitness': 'predicted_binding_fitness'})\n",
    "predicted_abundance = predicted_abundance.rename(columns={'fitness': 'stability_fitness'})\n",
    "predicted_binding = predicted_binding.rename(columns={'fitness': 'binding_fitness'})\n",
    "predicted_abundance = predicted_abundance.rename(columns={'mean_ddG': 'folding_ddG'})\n",
    "predicted_binding = predicted_binding.rename(columns={'mean_ddG': 'binding_ddG'})\n",
    "predicted_stability_vs_binding = predicted_binding.merge(predicted_abundance, left_on='aa_seq', right_on='aa_seq')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "b1788682-96c9-469e-80bd-0209905d2746",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x170 with 7 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%config InlineBackend.print_figure_kwargs = {'bbox_inches':'tight'}\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import gaussian_kde\n",
    "from scipy.stats import pearsonr\n",
    "import math\n",
    "from scipy.interpolate import UnivariateSpline, InterpolatedUnivariateSpline\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "\n",
    "xlim=(min(predicted_stability_vs_binding['stability_fitness']),max(predicted_stability_vs_binding['stability_fitness']))\n",
    "ylim=(min(predicted_stability_vs_binding['binding_fitness']),max(predicted_stability_vs_binding['binding_fitness']))\n",
    "\n",
    "predicted_abundance=predicted_abundance.sort_values('stability_fitness',ascending=True)\n",
    "x=predicted_abundance['stability_fitness'].iloc[35:]\n",
    "y=predicted_abundance['binding_fitness_from_abundance_fitness'].iloc[35:]\n",
    "y_spl=UnivariateSpline(x,y,s=0.0001)\n",
    "l=len(x.index)-1\n",
    "x_range = np.linspace(x[0],x[l-1],100000)\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=6, figsize=(7,1.7))\n",
    "coliterator=0\n",
    "rowiterator=0\n",
    "for hamm_dist in range(2,8,1):\n",
    "    \n",
    "    y = predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist][['binding_fitness']].to_numpy()\n",
    "    x = predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist][['stability_fitness']].to_numpy()\n",
    "    \n",
    "    model = LinearRegression()\n",
    "    model.fit(x,y)\n",
    "    \n",
    "    y_pred = model.predict(x)\n",
    "    \n",
    "    \n",
    "    ax[coliterator].set(xlim=xlim,ylim=ylim)\n",
    "    hbin=ax[coliterator].hexbin(x,y,gridsize=50,bins='log', vmax=50, rasterized=True,linewidths=0, extent=[min(predicted_stability_vs_binding['stability_fitness']),max(predicted_stability_vs_binding['stability_fitness']),min(predicted_stability_vs_binding['binding_fitness']),max(predicted_stability_vs_binding['binding_fitness'])])\n",
    "    ax[coliterator].plot(x_range,y_spl(x_range),'-', color='red')\n",
    "\n",
    "    ax[coliterator].set_title(str(hamm_dist)+'\\nr = '+str(round(pearsonr(x.ravel(),y.ravel())[0],2)), fontsize=7)\n",
    "    ax[coliterator].axline((0, 0), slope=1,ls='dashed',color='k',lw=0.5)\n",
    "    ax[coliterator].set_ylim(min(predicted_stability_vs_binding['binding_fitness']),max(predicted_stability_vs_binding['binding_fitness']))\n",
    "    ax[coliterator].set_xlim(min(predicted_stability_vs_binding['stability_fitness']),max(predicted_stability_vs_binding['stability_fitness']))\n",
    "    ax[coliterator].spines[['right','top']].set_visible(False)\n",
    "    ax[coliterator].set_xticks([0,-0.5,-1])\n",
    "    #ax[coliterator].set_xticklabels([0,-0.5,-1])\n",
    "    ax[coliterator].tick_params(axis='x', labelsize=8)\n",
    "    ax[coliterator].tick_params(axis='y', labelsize=8)\n",
    "    ax[coliterator].axhline(0, ls='--', color='k',lw=0.5)\n",
    "    ax[coliterator].axvline(0, ls='--', color='k',lw=0.5)\n",
    "    ax[coliterator].set_rasterization_zorder(0)\n",
    "    if coliterator != 0:\n",
    "        ax[coliterator].yaxis.set_ticklabels([])\n",
    "    coliterator+=1\n",
    "    \n",
    "\n",
    "fmt = '%1.f'\n",
    "cax=fig.add_axes([0.1,0.22,0.045,0.05])\n",
    "cb=fig.colorbar(hbin, ax=ax, shrink=0.7,format=fmt,location='top',cax=cax)\n",
    "cb.outline.set_visible(False)\n",
    "cb.ax.tick_params(axis='both', which='both',direction='in', color='white',width=0.25)\n",
    "cb.ax.xaxis.set_tick_params(pad=0)\n",
    "cb.set_ticks([1,10,50])\n",
    "cb.set_ticklabels([1,10,50], fontsize=6, rotation=0)\n",
    "ax1 = cb.ax\n",
    "ax1.text(60,0.1,'count',rotation=0, fontsize=7)\n",
    "\n",
    "fig.text(0.5, -0.015, 'Observed Abundance fitness', ha='center', fontsize=7)\n",
    "fig.text(-0.005, 0.5, 'Observed Binding fitness', va='center', rotation='vertical', fontsize=7)\n",
    "fig.tight_layout()\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_binding_vs_stability_observed_fitness_with_bindingfitnesspredictedfromabundancefitness_2ndorder3statemodel.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8194ab67-321f-4fcd-b983-5aae97f0a092",
   "metadata": {},
   "source": [
    "## Panel q"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "39a06ac4-238d-4554-b455-84fee8cc3388",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "predicted_phenotypes = pd.read_csv(path+'Fig7'+'/FYN_coreDTS_bPCA_order2_sparseregularized/task_4/predictions/predicted_phenotypes_all.txt',\n",
    "                                  sep='\\t', lineterminator='\\n', low_memory=False)\n",
    "\n",
    "predicted_abundance = predicted_phenotypes.loc[predicted_phenotypes['Abundance']==1]\n",
    "predicted_binding = predicted_phenotypes.loc[predicted_phenotypes['Binding']==1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "594d6339-cc9e-4cb6-b742-d5191817fba6",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Normalize data\n",
    "\n",
    "from seaborn import kdeplot\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "RT=303*0.001987\n",
    "\n",
    "predicted_abundance_scaled = predicted_abundance.copy()\n",
    "predicted_binding_scaled = predicted_binding.copy()\n",
    "\n",
    "def normalize(values, bounds):\n",
    "    return [bounds['desired']['lower'] + (x - bounds['actual']['lower']) * (bounds['desired']['upper'] - bounds['desired']['lower']) / (bounds['actual']['upper'] - bounds['actual']['lower']) for x in values]\n",
    "\n",
    "# Normalize abundance fitness measurements\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_abundance, x='fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "\n",
    "predicted_abundance_scaled['fitness']=        normalize(list(predicted_abundance_scaled['fitness']),{'actual': {'lower': mode, 'upper':         0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_abundance_scaled['fitness1_uncorr']=normalize(list(predicted_abundance_scaled['fitness1_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_abundance_scaled['fitness2_uncorr']=normalize(list(predicted_abundance_scaled['fitness2_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Normalize abundance fitness predictions\n",
    "\n",
    "for index, row in predicted_abundance_scaled.loc[predicted_abundance_scaled['WT']!=True].iterrows():\n",
    "    predicted_abundance_scaled.at[index,'predicted_fitness']=row['fold_'+str(int(row['Fold']))]\n",
    "    predicted_abundance_scaled.at[index,'predicted_ddG']=row['fold_'+str(int(row['Fold']))+'_additive_trait0']*RT\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_abundance_scaled, x='predicted_fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "predicted_abundance_scaled['predicted_fitness']=           normalize(list(predicted_abundance_scaled['predicted_fitness']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Obtain mean inferred folding ddGs\n",
    "\n",
    "predicted_abundance_scaled['mean_additive_trait']=predicted_abundance_scaled[['fold_'+str(i)+'_additive_trait0' for i in range(1,11,1)]].mean(axis=1)\n",
    "predicted_abundance_scaled['std_additive_trait']=predicted_abundance_scaled[['fold_'+str(i)+'_additive_trait0' for i in range(1,11,1)]].std(axis=1)\n",
    "\n",
    "predicted_abundance_scaled['mean_ddG']= predicted_abundance_scaled['mean_additive_trait']*RT\n",
    "predicted_abundance_scaled['std_ddG']=  predicted_abundance_scaled['std_additive_trait']*RT\n",
    "\n",
    "# Normalize binding fitness measurements\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_binding, x='fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "\n",
    "predicted_binding_scaled['fitness']=        normalize(list(predicted_binding_scaled['fitness']),{'actual': {'lower': mode, 'upper':         0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_binding_scaled['fitness1_uncorr']=normalize(list(predicted_binding_scaled['fitness1_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_binding_scaled['fitness2_uncorr']=normalize(list(predicted_binding_scaled['fitness2_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "predicted_binding_scaled['fitness3_uncorr']=normalize(list(predicted_binding_scaled['fitness3_uncorr']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Normalize binding fitness predictions\n",
    "\n",
    "for index, row in predicted_binding_scaled.loc[predicted_binding_scaled['WT']!=True].iterrows():\n",
    "    predicted_binding_scaled.at[index,'predicted_fitness']=row['fold_'+str(int(row['Fold']))]\n",
    "    predicted_binding_scaled.at[index,'predicted_ddG']=row['fold_'+str(int(row['Fold']))+'_additive_trait0']*RT\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "\n",
    "my_kde = kdeplot(data=predicted_binding_scaled, x='predicted_fitness', ax=ax)\n",
    "\n",
    "lines = my_kde.get_lines()\n",
    "plt.close()\n",
    "for line in lines:\n",
    "    x, y = line.get_data()\n",
    "    mode=(x[np.argmax(y)])\n",
    "\n",
    "predicted_binding_scaled['predicted_fitness']=           normalize(list(predicted_binding_scaled['predicted_fitness']),{'actual': {'lower': mode, 'upper': 0}, 'desired': {'lower': -1, 'upper': 0}})\n",
    "\n",
    "# Obtain mean inferred binding ddGs\n",
    "\n",
    "predicted_binding_scaled['mean_additive_trait']=predicted_binding_scaled[['fold_'+str(i)+'_additive_trait1' for i in range(1,11,1)]].mean(axis=1)\n",
    "predicted_binding_scaled['std_additive_trait']=predicted_binding_scaled[['fold_'+str(i)+'_additive_trait1' for i in range(1,11,1)]].std(axis=1)\n",
    "\n",
    "predicted_binding_scaled['mean_ddG']= predicted_binding_scaled['mean_additive_trait']*RT\n",
    "predicted_binding_scaled['std_ddG']=  predicted_binding_scaled['std_additive_trait']*RT"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "bd5d71c2-516d-4a06-9055-1843c429d1ff",
   "metadata": {},
   "outputs": [],
   "source": [
    "predicted_abundance_scaled = predicted_abundance_scaled.rename(columns={'predicted_fitness': 'predicted_stability_fitness'})\n",
    "predicted_binding_scaled = predicted_binding_scaled.rename(columns={'predicted_fitness': 'predicted_binding_fitness'})\n",
    "predicted_abundance_scaled = predicted_abundance_scaled.rename(columns={'fitness': 'stability_fitness'})\n",
    "predicted_binding_scaled = predicted_binding_scaled.rename(columns={'fitness': 'binding_fitness'})\n",
    "predicted_abundance_scaled = predicted_abundance_scaled.rename(columns={'mean_ddG': 'folding_ddG'})\n",
    "predicted_binding_scaled = predicted_binding_scaled.rename(columns={'mean_ddG': 'binding_ddG'})\n",
    "predicted_stability_vs_binding = predicted_binding_scaled.merge(predicted_abundance_scaled, left_on='aa_seq', right_on='aa_seq')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "602daa56-7501-4699-8517-8c1873c54679",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x170 with 7 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import gaussian_kde\n",
    "from scipy.stats import pearsonr\n",
    "import math\n",
    "\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "\n",
    "xlim=(min(predicted_stability_vs_binding['predicted_stability_fitness']),max(predicted_stability_vs_binding['predicted_stability_fitness']))\n",
    "ylim=(min(predicted_stability_vs_binding['predicted_binding_fitness']),max(predicted_stability_vs_binding['predicted_binding_fitness']))\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=6, figsize=(7,1.7))\n",
    "coliterator=0\n",
    "rowiterator=0\n",
    "for hamm_dist in range(2,8,1):\n",
    "    #predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist].plot.scatter(y='stability_fitness', x='binding_fitness')\n",
    "    \n",
    "    y = predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist][['predicted_binding_fitness']].to_numpy()\n",
    "    x = predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist][['predicted_stability_fitness']].to_numpy()\n",
    "    \n",
    "    model = LinearRegression()\n",
    "    model.fit(x,y)\n",
    "    \n",
    "    y_pred = model.predict(x)\n",
    "    \n",
    "    ax[coliterator].set(xlim=xlim,ylim=ylim)\n",
    "    hbin=ax[coliterator].hexbin(x,y,gridsize=50,bins='log', vmax=50, rasterized=True,linewidths=0, extent=[min(predicted_stability_vs_binding['predicted_stability_fitness']),max(predicted_stability_vs_binding['predicted_stability_fitness']),min(predicted_stability_vs_binding['predicted_binding_fitness']),max(predicted_stability_vs_binding['predicted_binding_fitness'])])\n",
    "    ax[coliterator].set_title(str(hamm_dist)+'\\nr = '+str(round(pearsonr(x.ravel(),y.ravel())[0],2)), fontsize=7)\n",
    "    ax[coliterator].axline((0, 0), slope=1,ls='dashed',color='k',lw=0.5)\n",
    "    ax[coliterator].set_ylim(min(predicted_stability_vs_binding['predicted_binding_fitness']),max(predicted_stability_vs_binding['predicted_binding_fitness']))\n",
    "    ax[coliterator].set_xlim(min(predicted_stability_vs_binding['predicted_stability_fitness']),max(predicted_stability_vs_binding['predicted_stability_fitness']))\n",
    "    ax[coliterator].spines[['right','top']].set_visible(False)\n",
    "    ax[coliterator].set_xticks([0,-0.5,-1])\n",
    "    #ax[coliterator].set_xticklabels([0,-0.5,-1])\n",
    "    ax[coliterator].tick_params(axis='x', labelsize=8)\n",
    "    ax[coliterator].tick_params(axis='y', labelsize=8)\n",
    "    ax[coliterator].axhline(0, ls='--', color='k',lw=0.5)\n",
    "    ax[coliterator].axvline(0, ls='--', color='k',lw=0.5)\n",
    "    ax[coliterator].set_rasterization_zorder(0)\n",
    "    if coliterator != 0:\n",
    "        ax[coliterator].yaxis.set_ticklabels([])\n",
    "    coliterator+=1\n",
    "    \n",
    "#fmt = '%1.f'\n",
    "#cax=fig.add_axes([0.99,0.25,0.005,0.3])\n",
    "#cb=fig.colorbar(hbin, shrink=0.7,format=fmt,cax=cax)\n",
    "#cb.ax.set_title('Counts',fontsize=7)\n",
    "##cb.set_label('counts', rotation=270, fontsize=7, labelpad=8)\n",
    "#cb.set_ticks([1,5,10,50])\n",
    "#cb.set_ticklabels([1,5,10,50], fontsize=6)\n",
    "\n",
    "fmt = '%1.f'\n",
    "cax=fig.add_axes([0.1,0.22,0.045,0.05])\n",
    "cb=fig.colorbar(hbin, ax=ax, shrink=0.7,format=fmt,location='top',cax=cax)\n",
    "cb.outline.set_visible(False)\n",
    "cb.ax.tick_params(axis='both', which='both',direction='in', color='white',width=0.25)\n",
    "cb.ax.xaxis.set_tick_params(pad=0)\n",
    "#cb.ax.set_title('Count', fontsize=7)\n",
    "#cb.set_label('count', rotation=0, fontsize=7, labelpad=2)\n",
    "#cb.ax.xaxis.set_label_position('right')\n",
    "cb.set_ticks([1,10,50])\n",
    "cb.set_ticklabels([1,10,50], fontsize=6, rotation=0)\n",
    "ax1 = cb.ax\n",
    "ax1.text(60,0.1,'count',rotation=0, fontsize=7)\n",
    "\n",
    "fig.text(0.5, -0.015, 'Predicted Abundance fitness', ha='center', fontsize=7)\n",
    "fig.text(-0.005, 0.5, 'Predicted Binding fitness', va='center', rotation='vertical', fontsize=7)\n",
    "fig.tight_layout()\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_binding_vs_stability_predicted_fitness_2ndordersprsregmodel.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe533674-1c61-48bd-93c5-416553201c93",
   "metadata": {},
   "source": [
    "## Panel r"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "ab7c489b-4e4b-4d98-b91c-e10d58e79499",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsQAAACzCAYAAACO7QrTAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAADU30lEQVR4nOzdd3gU1f7H8fduspveeyeBhBBK6L13pQhKE0ERRRFBQUGkeEHBiqJIUaSqNAEp0nsNBAghvZLee0+2z+8PfnIvF71SkmwC83qePJLdzcx35uPMnp09c45EEAQBkUgkEolEIpHoKSXVdwEikUgkEolEIpE+iQ1ikUgkEolEItFTTWwQi0QikUgkEomeamKDWCQSiUQikUj0VBMbxCKRSCQSiUSip5rYIBaJRCKRSCQSPdXEBrFIJBKJRCKR6KkmNohFIpFIJBKJRE81sUEsEolEIpFIJHqqiQ3ieqZUKhk3bhy9e/ema9euBAcH67sk0QPSaDS8+OKL9OzZk2eeeYbi4mJ9lyR6BHl5edjY2Oi7DNEjaN26NX379qVv374sXrxY3+WIHtLy5cvp2bMnnTp14siRI/ouR/SAVq5cefe4a926NW3bttV3SXVCbBDXs61bt+Ln58fFixf5+eefmTNnjr5LEj2gvXv34uTkxOXLl5kwYQLfffedvksSPYL58+ejUqn0XYboIZWXl2Nra8v58+c5f/48y5cv13dJoodw7tw5wsPDuXz5MkeOHCEhIUHfJYke0Hvvvcf58+c5d+4cTk5O/PTTT/ouqU4Y6ruAp83EiRORSCQAaLVajIyM9FyR6EFNmDCBMWPGAJCeno6lpaWeKxI9rLNnz2Jvb4+Dg4O+SxE9pFu3blFUVET//v0xNjZm1apV+Pr66rss0QM6efIknp6ejBgxApVKxerVq/Vdkugh7dq1i4CAADp37qzvUuqE2CCuZxYWFgAUFhby8ssv8+233+q5ItHDMDQ0ZNiwYdy4cYPTp0/ruxzRQ1CpVHz66accPHiQvXv36rsc0UOytLRk7ty5vPLKK1y5coUpU6YQFBSk77JED6igoID8/Hz279/PrVu3eO2117h06ZK+yxI9hG+++YY//vhD32XUGbFBrAeJiYm88MILLF++nF69eum7HNFDOnLkCLdv32b48OHExcXpuxzRA/riiy+YPn065ubm+i5F9AhatGhBQEAAEomEHj16kJ2dre+SRA/Bzs6Odu3aYWhoSKdOncjIyNB3SaKHEBERgaurK66urvoupc6IfYjrWVZWFiNHjmTDhg2MHDlS3+WIHsKGDRv4/vvvgTtX+qVS8fBpTE6fPs3atWvp27cvubm5DB8+XN8liR7Cjz/+yEcffQTc6T7h6emp54pED6NHjx4cP34cuHNRSOy21LicPXuWoUOH6ruMOiURBEHQdxFPkxkzZvDHH3/QrFkzABwcHNizZ4+eqxI9iPLyciZPnkxpaSk6nY7PPvtMvMLfSDVp0oTU1FR9lyF6CDU1NUyaNImCggIMDQ354YcfaN68ub7LEj0gQRB4//33CQoKQhAE1q5dS6dOnfRdlugBvf3224wYMeKJbhSLDWKRSCQSiUQi0VNN/M5XJBKJRCKRSPRUExvEIpFIJBKJRKKnmtgg1pP4+Hh69OhBjx49+Pjjj+97Pi8vjwEDBtCrVy/eeustdDodgiDw9ttv06NHDzp16vRED3/SUO3bt49OnTrRtWvXv5xpKSgoiM6dO9OtWzc2bNgA3Blv2sXF5e5MP2vWrKnvskX/71Hy++CDD+5m5+vry6hRo+q5atGfHiW/oqIihg4dSt++fendu7c4IUQD8U/vgX+aM2cOW7durb/CRH/rUTN755136Nq1a8M//gSRXgwfPly4du2aoNPphGeeeUaIiIi45/mZM2cKv/32myAIgjB9+nThjz/+EI4fPy689NJLgiAIQmFhodC0adN6r/tpplKpBH9/f6GiokIoKysTAgMDBZVKdc9r2rdvL2RkZAhKpVLo1KmTUFBQIERGRgqTJk3SU9WiPz1qfn+qqakROnbsKKSmptZ36SLh0fP78MMPhQ0bNgiCIAjHjx8XJkyYoI/yRf/ln94DS0pKhOHDhws+Pj7Cli1b9FOk6B6Pkll4eLjQt29fQRAE4dSpU8KLL75Y32U/MPEK8UPaunUrPXv2pHv37ty8efPu47/99tvdq0h//mRlZf3tcqKjo+ncuTMSiYQhQ4Zw9uzZe54PCgpi8ODBAAwdOpSzZ8/Su3dv1q5dC4BEIsHQUBxG+lE8aoaxsbE0b94cc3NzLC0t8fb2JiYm5u7z5eXl6HQ63N3dkcvl9OjRg6CgIG7evEl8fDx9+vRh7Nix5OXl1ev2PmnqO78/ffvtt7z44ot4eXnVz4Y+oeo7vw8++IBJkyYB4uygtaG+3gOrqqpYsGABkydPrrNteVroMzMvLy/kcjlqtZqKioq7k5M1RGKL6hG4ubnx22+/3fPY+PHjGT9+/CMtz8LC4r5B5svLy7Gysrr7fEVFBSYmJpiYmFBdXc2ECRNYsmTJo22A6JEy/M9M4N+5/Ofz/zmd85/Pe3h4sHDhQkaNGsWOHTuYM2cOO3bsqMWtefrUZ34ACoWCbdu2ERISUlub8FSrz/xsbGwASE5OZt68eezfv7+2NuOpVR/vgW5ubri5uXHy5MlHrlP0b/rKTBAEVCoVzZs3p6ysrEF39RQbxI/gr8a+/O233/jhhx/ueWz79u24ubkBcPnyZRYvXgzA8uXLkUgkd19XUVFxz4ke7kxTWlFRcfe/fz6fn5/Pc889x5QpU3jxxRdrdbueJo+S4Z9Z/Om/c/u757t27YpMJgNg1KhRfPrpp7W6LU+j+swP4OTJkwwYMAATE5Na3Y6nVX3nd+3aNV599VU2b96Mn59frW7L06g+3gNFtUtfmf3yyy/4+Phw5syZu/dGhYaGYmxs/DibUyfEBvEj+KsZyv7pk1bPnj05f/783d/9/f0JCQmhQ4cOnDhxgmXLlt3z+m7dunHy5EnGjBnDsWPHGDRoEFVVVQwaNIhly5aJs9w9pkfJ0N/fn7i4OCoqKhAEgfj4ePz9/e8+/+fVqczMTBwdHbl06RLz589n4cKF+Pn5MWPGDE6dOkX79u1rf4OeMvWZHzwdszTVp/rMLzw8nFdffZWDBw/i6+tb+xvzFKqP90BR7dJXZtbW1lhZWSGVSrGxsUGtVqPRaB5pG+qa2IdYT7755hvmzJlD586dadeuHR06dKCgoICXXnoJgI8++oj169fTrVs3lEolo0ePZu3atWRlZbFy5cq7/X20Wq2et+TpIZfL+fTTTxkwYAB9+/Zl+fLlyGQyTp06xYoVKwBYvXo1L7zwAt26dWPq1Kk4OzuzcOFCDhw4QN++ffnpp5/uvlZUvx41P7gz1ayPj48+y3/qPWp+CxYsQKVSMW3aNPr27cusWbP0vCUi+Of3QFHD8yiZTZw4kfLycnr06EHfvn3517/+hbm5eT1W/eDEmepEIpFIJBKJRE818QqxSCQSiUQikeipJjaIRSKRSCQSiURPNbFBLBKJRCKRSCR6qokNYpFIJBKJRCLRU63BNIgXLFhAixYtCAgIYOXKlfouRyQSiUQikUj0lGgQ4xAfPXqUoKAgIiMjUalUtGzZkmHDhv3lQNIikUgkEolEIlFtahAN4meffZZBgwZhaGhIVlYWWq0WMzOz+16nVCpRKpV3f/9zSkB7e/t7ZlARNUxifo2bmF/jJWbXuIn5NW5ifo1Dg+kyIZPJWLx4MS1atKB///53pw78T59//jlWVlZ3f6ytrXF0dLxnus66UlZWVufrqAvFxcW88sor+i4D0F9+jTU7gHPnzjWYLkRifg9vwYIFREVF6bsM8dz5CFQqFePHj28Qs2qJx97Di4uLY+7cufouAxDzexTr16/n8OHD9brOBtMghjtzZRcWFpKZmcmGDRvue37BggWUlZXd/cnIyKi32iZPnlxv66pNERERvP322/ouA9Bffo01O4DY2FjeeOMNfZcBiPk9LLVajaOjI61atdJ3KeK58xEkJSUxadIkDA31/0WqeOw9vNDQUGbPnq3vMgAxv0dRUVHBs88+W6/r1P+RDkRHR6PT6WjdujWmpqaMHj2aiIiI+15nZGSEkZGRHipsnPbs2UPbtm3x9fXVdymAmN/DEASBr776irlz52JgYKDvcgAxv4dRUlLCrl27mDNnjr5LAcTsHtbFixeRSqWMGDFC36UAYn4Pa82aNbz88stYWlrquxRAzO9hqFQqVq1axbx58+p93Q3iCnFsbCzTp09HpVKhVCrZt28fvXv31ndZ92hs86unpKSwZcsWfHx89F2K3jW27AC2bduGSqVqMI1hfWps+QmCwDvvvEP79u31XUqD0NjyKy0tZenSpQQGBuq7FL1rbNnBnQ8zYWFhDaYxrE+NMb+PP/6YJk2a6GXdDeIK8ZgxYwgNDaVt27YYGBgwbtw4xo0bp++y7mFlZaXvEh7KjRs3WLt2rdigovFlB5Cfn8+CBQv0XUaD0NjyKy0tpVevXnTp0kXfpTQIjS2/8PBwPvvsMywsLPRdit41tuzgzjfO3377rb7LaBAaW34ajQZbW1vGjh2rl/U3iCvEAJ999hkxMTFERkby0Ucf6buc+6xbt07fJTywn376iQEDBuDt7a3vUhqExpSdRqNhxYoVvP/++w2i72JD0JjyS01N5fjx4w2m33dD0Jjy+/3333FxcaFr1676LqVBaEzZAaxYsYI33nhD/DDz/xpTfmVlZfz000+8//77equhwTSIRbXj2rVrXLx4EVtbW32XInoEX3zxBY6OjvouQ/QINBoNM2bMEK8MN1Kpqals2LBB7GbWSG3fvp3KykrxW9FG6p133qFt27Z6rUFsED+gNWvW6LuEBxIWFsbq1avFsQ3/Q2PJTqfTIZfLefnll/VdSoPSWPJLS0tjypQpYoPqvzSW/K5fv87atWvFb2b+Q2PJDiAnJ4fFixfru4wGpbHkV1paSrdu3ejevbte6xAbxA9o06ZN+i7hH61cuZKpU6diY2Oj71IalMaQXWVlJevWreODDz4QP8z8l8aQ340bN0hNTW1w9z40BI0hv40bN9K/f3+aNm2q71IalMaQnVar5euvv2bu3LnIZDJ9l9OgNIb80tPTOXbsGNOnT9d3KWKD+EHdunVL3yX8T/v37yc3N1c8IfyFhp4dwPvvv09AQIC+y2iQGnp+VVVVzJ8/n3bt2um7lAapoecXEhLC2bNnsbOz03cpDU5Dzw7u9BsWuwj+tYaen1arZcaMGXTq1EnfpQBig/iB/dXMeQ1JcnIyy5Yt03cZDVJDz66yspKWLVvSv39/fZfSIDX0/CIjI1m8eLH4pvw3Gnp+N2/eFLuZ/Y2Gnp1Op0MikfDqq6/qu5QGqaHnl5aWxqRJk2jWrJm+SwFAIgiCoO8iHlV5eTlWVlaUlZXV+ZiDCoUCY2PjOl3Ho9DpdHzzzTfMnTu30Z3Q6yu/hpod3On3dvr06UY5o5CYHxw7dgx3d3dat26t71IeinjuvOO7775jxowZyOVyfZfyUMRjD6qrq9m8eTMzZ87UdykPTczvzkyCxcXFDBw4UN+l3CVeIX5ADbVv4OrVqzExMWl0jeH61FCz0+l0zJgxQ5zA4R801Pxyc3NZuXIlfn5++i6lQWuo+f3xxx9kZGQ0usZwfWqo2QHMmzeP5s2b67uMBq2h5lddXc28efP0PqrEfxMbxI2YIAhUVVXx9ttv67sU0SPIyspi9OjRtGzZUt+liB5BcHAw3333nTglayOVmJjIZ599pu8yRI+gqqoKX19fBg0apO9SRI8gMjKSBQsWYG9vr+9S7iGOL/OAxowZo+8S7qFUKvnpp59YuHChvktp8BpadnBnNqX09HRxiLUH0BDz2759O/3798fFxUXfpTR4DS0/QRBYuXIl7733nvjN2j9oaNkB5OXlcfLkSWbPnq3vUhq8hpjfiRMncHFxaZDjtYtXiB+Qp6envku4x6JFi/Q233dj09CyUyqVzJ49WxyV4AE1tPxiYmL4/fffcXZ21ncpjUJDy+/PsYbFxvA/a2jZCYLAjBkzGtxX7Q1VQ8svLy+PFStWNNhuZmKD+AGtXLlS3yXcpVAocHJyYsSIEfoupVFoSNnBnQbV7NmzxQbVA2po+V29epW1a9eKDaoH1JDyEwSBsrIyZs2ape9SGoWGlB3c6WY2fPjwRncTq740tPyCg4NZuXJlg73RT+wy0cgUFRXxxx9/MG/ePH2XInoE58+fx8LCgmHDhum7FNEjWLduHVOmTMHU1FTfpYge0p/dzBYtWqTvUkSPIDY2luTkZHGItUZq586d9OnTB1dXV32X8rfEK8QP6JtvvtF3CQiCwMyZM8VPxw+pIWQHUFxczCeffIK/v7++S2lUGkp+Z86cITw8XGwMP6SGkt9HH32Eu7u7vstoVBpKdiqVinfeeUfsZvaQGkp+sbGx7Nq1q8F/K9pgGsQrV66kZcuWtG7dmqlTp6JSqfRd0j327t2r7xLIz8+nb9++dOzYUd+lNCoNITuAa9eu8eWXX2JmZqbvUhqVhpJfeHh4g3mDaUwaQn5KpRI7OztGjx6t71IalYaQHdzpZjZr1qwGfXWxIWoo+V25coV169YhlTaYJudfahDVXb9+nS1btnD9+nUiIiJQq9WsXbtW32Xd4+rVq3pdf1JSEsHBwbz55pt6raMx0nd2AL///jsBAQENZorKxkTf+QmCwLfffsvs2bMxNzfXay2Nkb7zKykpYceOHcyfP1+vdTRG+s4O4OLFi2g0GkaOHKnvUhqdhpDf+vXrmTBhQoOfNQ8aSIPYxsaGNWvWYGZmhkQiITAwkPT0dH2XdQ99znOv0Wh4++23CQwM1FsNjZk+s4M7H2Y2bdokfl37iPSd3y+//EJVVVWDv7rRUOkzvz+7mQUEBOithsZM38deSUkJS5cuFbuZPSJ953f+/Hlu3LjRaL4VfaCpm4uLi/nyyy/58ssv6dWrF5mZmUilUg4dOlTrJ5r8/Hw6d+7Mli1b6Nev3z3PKZVKlErl3d/Ly8vx8PCol+lH9Sk6OpqYmBjGjh2r71Iey9Oa388//0zv3r3x9vbWdymP5WnNb/ny5SxYsAADAwN9l/LIntbsCgoK2LNnDzNmzNB3KY/lac3v+PHjWFtb07VrV32X8lie1vxWrlzJtGnTsLCw0HcpD0Z4ABMmTBC2bNkiCIIg9O3bVxAEQbhw4YIwatSoB/nzB5aSkiIEBAQIy5cv/8vnlyxZIgD3/ZSVldVqHX9lxIgRdb6OvxIcHCxcuXJFL+uubfrKT1/ZCYIgbNq0qV7+/6wPT1t+arVaWL16tV7WXduexnNncnKysH//fr2su7Y9bceeIAjC/v37hZSUFL2tvzY9jfmtWrVK0Gq1elv/o3ig7wBzc3OZMmXKPY/17t2bnJycR2uF/4WwsDB69OjB9OnT/3ZYnAULFlBWVnb3JyMjo9bW3xBVVFSwYMECWrRooe9SasXTll9wcPDdYdaeBE9bfp9//jlWVlb6LqNWPG3ZPWndzBpTfhUllY+9jNTUVH766acnpptZY8qvNmzfvp2SkpJG183sgcYhVqvVd/99+PDhu/+ureGHCgoKGDp0KOvWreP555//29cZGRlhZGRUK+t8WMOHD6/3dYaEhLBkyRKsra3rfd11QV/56SM7uDOqxKpVq56YCRyepvw0Gg0SiYRJkybV+7rrwtN27kxMTOSVV15p9N2U/tRYjr1zu4L4/KVVzN08g8Gv9H3k9V68eJE1a9ZgaPhkTJXQWPKrLSkpKY1yvO8Har7b2dkRGRkJcLdzdFRUFDY2NrVSxHfffUd5eTmffPIJbdu2pW3btg1uZ7Zq1ape13fkyBHc3d3p06dPva73SVTf2QGsWbOGt99+u9aOkadZfedXWVnJzz//zOLFi5+YDzP6VN/53bhxg5KSEsaPH1+v630SPWx2nZ9py/xfZtFt5KMPDbp161ZGjhyJj4/PIy9DdMfD5Hfox5O82W4uOSl5j7w+rVbLmjVrWLx4caP8MPNADeIVK1bw0ksvsWjRIjZt2sTChQuZNGkSK1asqJUiPv30U6qrqwkLC7v78+mnn9bKsmvLF198UW/rysrKYtWqVXh5edXbOp9k9ZkdwL59+8jIyGiUJ4SGqL7ze++998Q341pUn/lVVlYyf/58cVSCWvKw2ZlZmTHgpV5Y2Dza8IQ3btzg5MmTT0xXJX37p/yyk3L5bvpP3A5Lwd3PhTa9AjCzfPRv/hv7OPsP1CD28/MjODiYNm3aUFhYSMuWLbl8+bL4plFHLl68yPfff49cLtd3KaJHEBsby7Jly/RdhugRlJWV4e/vf98INyL9y0zI5sKeq2i12r99TUhICIsXL8bW1rYeKxPVlqCgINasWSN+M1NPSgvKiQ1OoDCzmHb9W/P291OxtHu0e160Wi0ajea++80alUe5E6+mpkbYtGmT0Llz59q+ye+hlJWV1dud0hEREXW+DkEQhO3btwsFBQX1si59q6/86is7rVb7xIxK8CCetPyys7OFPXv21Mu69K0xnjtXvLpWeNbkRSEvLf8vnz927JgQHx9fK+tq6BrrsVeUUywoqhV/+dz3q74XclJza3V9DVVjze/vVFZWChs3bqyXddWlh7oFMDY2lnfeeQdfX18uXbpEQUFB7bfQG6iTJ0/W+ToiIiLYt2+f3gfTftLUR3YA33//faO7q7YxqI/8dDodb731ljiBQx2orfze3/QWu3M34ujpcN9zOTk5fPPNN2I3s1pWm8deTZWCl7zeYtGwz+977uDBg+xas5dXmr1DfkZhra3zaVdf733z5s17Io69B3r33rlzJ/369WPGjBm0b9+euLg4tmzZ8lQ13C5cuFDn67h48SJr164Vvy6qZfWRnU6no6SkhLfeeqvO1/W0qY/8UlNTee6558QGcR2orfwkEsnf9m88f/483333nd5G0nhS1eaxZ2JmzIqzS3l/0/3nyMjISHaf2MVXp/6Fo4d9ra3zaVdX584Dq48x0Ws6abGZlJWVEXP6Nie+CLrnNTdOhDHB/Q2uHgqpkxrqwgPd9fPWW2/x5ptvMmvWrHvGBXyaGm7m5o92k8CDWr9+PW+88YbYb7gO1HV2CoWCX3/9lY8//rhO1/O0quv8oqKiKCws5NVXX63T9Tyt6jq/Xbt2MXDgQBwc7r9yLHo8tZ1dqx733uwoCALr1q1j8eLFALh5u9Tq+p52/5RfWmwmv368m+dnDyegq98DL7dFV18Gv9IXFQqW/utnOrR7lr4DWt7zGo/mrgx6uS9eAY1nLOkHukKcnp6Oj48PY8eOZcCAAWzdupXKyscffLsx2bFjR50t++TJk0RERIiN4TpSl9kBLFy4EBcX8UReV+oyP4VCwezZs8VRCepQbed381Q453bduRoVHR3N7t27sbcXryrWhb/LLvRMJGd3XPrL5/IzCjmw+hiRl2I59ONJVArV3y5/7dq1//MmSdHj+V/HXlV5Nad+Pk9OSj6KSsXdx6Mux3Ly5/PodLr7/qa4qJL9e2/g2NSZVz4ez4wZM6hMrsHcx5Vhbwy657XOTRx57bOJuDZ1rr0NqmMP1CC2tLTkzTff5OrVq3z//fdERETQqlUrYmNj67q+BmPs2LF1tuyQkJBaG8JOdL+6zK66uhp7e3u9DYD+NKjL/MLCwpg9ezbOzo3npN3Y1HZ+u748wIb5vwJw7tw51q1b91R9W1mf/i673SsOsmH+tvseFwSBoxtOs/HDbWxbvpeNH26jMKuYosIK8vPKyUzMoaKkktvpBSiUavLz85k1a1Zdb8ZT66/yy7qdQ3lxBemxWRxYc5xWPfxp2bM5t2+lUFOtYvsXB9gwfxvKGhU5yXmUFZaTEpWOolrJyYMhbPjxLJERmaSmpmJZY0fp9Rw++HAYOSl5lBaU6WEra89DD5TasmVLVq5cyVdffcUff/xRFzU1SEqlstaX+efXRQsWLBBP6HWoLrIDKCoq4tSpUyxcuLBOli+6o67yO3/+PPb29nTt2rVOli+6ozby02q1GBgYAPDVqX+h0+rYuHEj06ZNE/sN16G/y+7zY4vQae+/gnhs81m2L/+dwD4tCT0VwRcnFuPa1JlJ49ZQWV5DzekbePQOIMLRBHejXPb9/O3dv/3PjP/Og7xG9G//nZ9KoWKq/7s079yM7698xuwf3+DLl1eTGplO6JlIAl4bQnS5jq3Xv0BuLGNK83fw8HMhLTaLVr38iboUx5RlL2LvqCM1NZXfg3eiUWswMDTgBYepeLZw56fwb/S0tY/vkW+JNzQ0/J/TLD9pBg0a9M8vekibN2+msrJSbAzXsbrIThAEZs6cia+vb60vW3SvusivqKiIZcuWPTFT+zZkj5vf4fWneN72VaKC4oA7966cv3Cemzdvio3hOvZ32UkkEgwM722YhsVn8vWVGNpP6UNCaBIzV0+lw6BAAFaufpl1m1/nq4vLKTa1RJZxlQnDe97921+W7mas42tkJmT/bS0Xdl9htM0Ugg/frIUtezr8d35yYzk/Ra7k4wPzAeg/sSerrnzKkv3z+P7qZ3y68iV+3PQa7p72GBgYsCFyJV+dWcKC7e+SHJ7OczOHMurdIbz77rt3u5kZygx5p9tCPPzd+Pz44nrfxtokjhH1gHr06FGryxMEgbS0NObOnVuryxXdr7azgzuzCfbq1YsOHTrU+rJF96qL/C5cuNDoZ1VqLB43v3YDWtFtZEd+WbqbgswiAIKDg/n6669rozzR/9CjRw9ighOYP2QZsdcSAdi78hAfj/kaZY0SlULFx2O+ZveKg3i72fHa890YP30Ikz8aS59x3dn//VGWPv8Vu7YHsf2Xy+w/GoGXpyW6/Eo0CTB/yDJ2fbGfm6fCad6pGT+8t5Wqsqq/rKVFNz8mfTQG3w7ihGAP6r+PPa1Gyy9Ld3P0p9MApESm8/OS30gMSaZFF1/MLUxo5uvMxg+38cXL3+Pp74bUWM7Ja6kEPNuehDI1K7/ZgYVZICrlvzsYjHl/JGZWphxccwy4c7Peh0OXc+NEGAAnfz7P4hGfU5JXWi/b/ageqsvE1KlT7/1jQ0Pc3d2ZMWPGE39Tw9KlS2uti4harWbLli188skntbI80f9Wm9kB3L59m+TkZGbMmFFryxT9vdrOb//+/XTt2hVXV9daW6bo7z1sfjWVNRxYfZzAfi0J6OqHRq1Fq9WhqFbyxw8nqHIsZOHCheI3a/Vg6dKlrPjXtxiZyDEwvHP9TGYkw8jk/28Al0gwMpUjM5JhZW7CS8M6AtC+sy8xwQmEnY+mVAuVaYXocgqpUtWQnRNDJ8fORF6OpVqlJSylCOH/rzbLTeTwN7k6etgzbt5zdb/RT5C5s+cx7ZmZjJo1FBNzEwCMTOXIjWUASA2kGJsa3c0W4PqxWySFp+Hgdme2R4lEgkxuiForkFeWRJPWbWkV0OWeMfdVChUSiYTE0BQu7Qsm6nIcGpX67rcIhnJD5CZyJNKGfcw+1BVinU5Hr169mD9/Pn369KG4uBgvLy9eeeWVuqrvibR8+XIsLB5tekSRfmk0GmbOnEnz5s31XYroESQlJbFx40acnJz0XYrob5QWlHNgzTGO/HSKmsoaYq8mEHYmEtemzmz4cQN5OfliY7iOxV5LpDi3BACfNp6MmvUs3q09AXju7aF8+Os7GJkYITeS8eEv7zBq1jOEh6VRVlpF6JlIysqqOLTtInHXErFs7klOThnFkalcurYLVVo1Lt7OJIamYOnlREJ2OZ6tvEiOSOP1z1/627Gm/5fbYSnkpT09E4U9qNL8O8dSSV4ZWbdzyIjP5oMtMxk7dyQRF2Owd7Pl4/0f0Kpni7t/c3HPVbISc5jx/VRuh6WgKK2iey8fYs6GcCv6KC9PfZZPPhuLh+e/56E4tvEMVVUKUmMyOfjjKc5sv4R7oA9N2/tw62wkfcZ241+738fawUofu+HBPcy0dv3797/n94EDBwqCIAi9e/eulWnzHlZ9Tj96/fr1WlmOQqEQlixZUivLauzqK7/ayk4QBCE0NFT47bffam15jVljyE9RrRB0Ot3d33/44QchJSXl7u81VX89jeyTTh/nzurKmnse/+/f/1PYuShhoGSMsO79n+8+ptPphA/mzRc0Gs29y6lW1mK1jUNd5leSXyoMNhgrfPDMcuH69evCkZ9OCQMlY4SjG08LgnBnf2s0GkFZ8+/9futWqjCg13LhvXHfCgMlY4R3P9gqdJn4tXAzJl3QanXCS+NWC0MHLhG+W7Xq7jF3bNMZYaBkjPDbVwfuLLeyRtDpdH85tfP/+n9Fq9UKQ+Xjhbc6zKvN3VCn9HHunNJ8ljDK5hVBEAQhKTxVGCgZI3z92jpBEP59DN3NQKESamqUwhCj8cKULh8KXSeuEAaPmyO0HPyOsHH3pbvL1Ol0Qk2VQlCr1EKvyd8IE97dKAzsvVx4560twoBey4W5Y1cKAyVjhNPbLtbpdtaWhx5l4tdff6Vr164EBwdjYGBAREQECoXin/+wkbt+/TqdOnV6rGWUl5dz6NAhli5dWjtFiR5IbWQHd/otyuVyxo0bVwtViR7Uo+ZXU1nDRM+3aNMngI/3f8Avv/zC5MmT7/Yb/uLl77l6MIStiauxcWzgVy4asevXr+No5syMjvMZNfMZpn01md++OsjWj3ay4sySe65O/Smwb0vm7JnHqjVn8DgYytBhbZjYfwo1t6BoVsnd2cxyskuZNuUnBg9twzvvDa3vTXsiWTtYseLK5yxYtJfkL7eyZ+9aWvVqgae/G6tWHuPU8Ui8K8vIuZ3D3vzN5BaWM+f7g7Ru40j0T2eYsGA0U5ZNICu/DE9nG+a/t4O8/GwKnCsJLvTkjPPrtO7VgtDTEXQc3JbNi3YSfTWekONhNO/UjNSodHZmrsfI5M4NkzHBCcztu4RJ/xrLxIX338gvlUr5NWUdJubG9b2rGrz/PHeuuvIpWvWdMZ992njxc+JqHD3tOXk8gq+/PEx3N3Nu7AvG/9UhJKUUYmFpjNPw7hSfDsFem0WFkTvDmrqze+Jqmuok9B3fg2XjVxJ6KgJbZ2u8dQJrb60gKjydjxf/TodO3ty6mUqH0V355rV1uPk649+5Yd+ELhEEQXjQF+fn5/PNN98QExND8+bNmT9/Pjdu3MDLy4uWLVv+8wL+h4qKCrp3786hQ4do0qTJA/1NeXk5VlZWlJWVYWlp+Vjr/ycjR4587H6MU6dOZcqUKfTu3buWqmrc6iu/2siuoqKCkSNHsn//fqytrWunsEauoeR39VAI25fvZfaPb9Ks3b9HjRAEgYt7ruLs40SxOp9169bxyy+/3P26PTkijdToDPqM6/bUDeVU3+fOvXt+5/yuIFr2aI5bMxfyMwoJPRVBn3Hd7vZt/GTcN5hZmPD+pjt98ysrFVw6H0eXbs1Ys3YlFnIrWru3o9+EHnf7Jmo0Ws6diaG5vwueXk/2fSz/qa7zU6u17NtxhQ/mvMam7zcycNKd96y01EIS4nNwNDGgsriSjkMCWTj8c4SmTpRGZeDkaU+WkwV93BwI23eNgO7NuXUzmbDqs8z5ZDk+TbzJOhfH0dMx2FsbkW6go0mpkvzYTBzaeFOk1PLCqPYMmNCDRcM/x9Xbkdthqbj7uTJl2QScvJ6M2Qjr4/g7/1sQr06fwvypi0m4mYSTlwM1lQqW/j6P/KIKFq0+xDM9W9K/YzN+XnWCyH1BOLjZUlSuwNLVjpryKnTuTmhy0zl+cRtvvLaECa/35/gPJ4i5Gk+7gW24eTIcC39PqtQ6pBIJ3oFepJ8IwbypG3mx6Th3aUF2bil923kw/t1nMTZt2KPCPNQVYiMjI9q3b09AQAAAx44d4+WXX37sIq5du8abb75JfHz8Yy+rrhgaPvTF9HsUFBQQEBAgNob14HGzA7h06RJLly4VG8N68E/5WTta4dXSA1NLk3sel0gk9BnXHYAvvtjD6tWr7+l76tPGC582XrVfsOgurVZLdmIuVw5cZ/Arfe8+XpJbSmp0BoYHDEiLycK+gy/lxiZUZRRxZMNpshNz8GnbhNQbSRiWl3Lt6E227t2Eg7s9x7ecozinBEEnYOVgSW5KHrZSHZ5e9iSFp3Lq5/M8P2f43avIf6qpUbH9lyDaBHrQuWuzu4/fOhtJyPEwJi56HjMrcdQRALVaQ15BBcZmJti62HA7LIXTv1zAd2BbUtOLUbpYU1qmorOBFE9/dzKVAlZejlQ1c8C8WknchWicvBxw8XEkODaWdn59kUaWkpscS25yHhZWRlR722NdWoU6uxxLewu0VTVYeDiTWa3j928Po1FqyErKRW4so7qihuLc0nsaxBf3XiU5Io1JH43BUPb45/gnjY2zNeaWZrg0dUJVo8LOzZawc1EcXn+KXhN74uNuj6OtOWWlNZQr1Fg5WaNVa7GQSZEUl2Gs00FJOZEZUQwJGI4mNY+spDzKLC0wsbfkdmgK5tZmSBUqTN0dkRhI8fC0R9fUmcKsAqxtzVEL4OXrwsDJfbgek0FMci6vjuqKkbxh5vVQVT333HP4+fnh5uYGUGs3Nvz444+sXr2ayZMn18ry6sK+ffse+W8zMzMJDw8Xh1jTk8fJDuD48eO0bNkSLy+x8aQP/5Rfiy6+tOjy91/FbdiwgQ8++OCeu6JF9UOr1tLJrC+nt12keedmuHjfuZkxLSaToAPXSY3OIDclH+v8amoECZqsIs7tvExBRhGKaiVXjl7nxPlj2JS6cWH3VTqO6sLZvcEUpxcgAbxaepASmUZhdgk+bbzIjM/m6uGb9Bnf474GsUKhJvhKIsbGhvc0iG/fSiX4yE2emzlUbBBz55uVK0dCCQlJ4eMvfqD9gNZc2H2FoIM3SKvUkFejwdvHkbSkPDr4O/HysgnMnb0dqxYuZBSX46uTkBWXTeDC0UicBSTONmiqrTi3Mwi5kwVl6UU4BXqRWlCKr9SA3IRsOj/TjuvHbuHoZEfQ5XiMEtJwcLYiN7UAGy9HkiPTObXvOh4BHphb3OkaER0UT8TFGCZ8OFpsEP8FrwB3fvhuPV2GtUcmvzOqRNjZSG6djWT4m4NY8PpgAIKvJpKQXICnpSkZYSk4NXGgsKAMIytjkiPO4yhxxc7TlaTIdFTHwkgrrKGJpRkZt5JxbeFOWVoeNk3dqKxU0KmHLz26NeWjEV/g0iOA1JwyBjaxBYWCE8dDSSitpLm/Mw7lauSmckpyy/AKcMe5iaM+d9VdD9Vlon///pw9e7bOimnSpAnnz5//2y4TSqXynplXysvL8fDwqJev/SZNmsS2bfdPVflPdDodo0aNYsWKFU/9yAT6yu9Rs4M7H2amTp3K4cOHkcvltVxZ49IY8/t16zYioyP4asVX9zxekleKjZN1LVTXONRXdiXFldjYmgOgqFai0+qY+NJLKE8Z03N0ZxbtmH33tV9P/4mTG0/TtK036TEZmFmaYmxtSkFaId2Gt+dfe+YysPtQSq7WMPTZIdw4FU7TVwaRmlyIVXI6RoZSfk5cQ+y1RGb3WMTw6YOZteb1e+pRKdWoalSYWzfuhm59HnvZaQW81mI2nYcEkmQWwferfsTWzpwN87ex95s/WHXlU3wCvZnaag46lZqt8d8TmpbD67/uo5e5E7cic+jY2pProYlk3tiOV6cXaXYrF6Vaze1XW+GHEcWfnqHl5D5cVdQw48VeTB7WiaXjv+X6oRAcn+9FRXEVO36fxe24HN57bwetW7gQHp3N2+8MZtQLHSnNL2tUx68+zp2/LN3N/I/nceLCMVp09UVRpcTC5s6xWVZajbmFMQYGdy4SbFywnd++OkjLAa2IuxSL5OtBpERFUbFhH0O7TSbD0BT/Tp5cz8qllyAlbNcVWk97hqj4XNasmYxHE3vGjVqFm6s1aza9zqXT0Xz2+SECva0I//UibQYFEnUuCptFQ0kpLcdtRzg29hbkJOXR84UuLNw+u072wcN6qEsmgYGBHDp0iLS0NNLT00lPT6+ruv7S559/jpWV1d0fDw+Pelt3eXn5I/1dXFwczz///FPfGAb95feo2QGcPHmS77///qlvDEPjy6+6ooaPp30BEfcO47T1X7sY7/oGkZdia6O8RqE+stuy8TzjRq8iKjIDgBkd5/NqwGyuXo3D/7UhfPjrrLuv/XXrJU7EFhDQpyWpUek4etqj0eoo8nDD95VBLNo1hyOHgknLgTmbPmH54QV0+nIikcoKXNxsqG7uzY9hdybmaNHFlw9+mcUf606w55tD99Q0r/9SXvR4E53u/mmGG5P6Ova0Wh1vvLEFn0kD+Nee90lLzWHc6FXs/e0a076cxK6sn/Dv7Mt77/5KqbcHlaVVzOi/lLmf7qNPsQlJiw7QycOGS9mZOJhVYuvbhzapFZTklLIx9Gt2vfkiMboaPD4ZSdTtEjq39OCbk0HMX7ybq/kKnN4dSqpchWFYPNPbzeWD7gtooakm8qdjjOniyegxnVg/9xcmuL1BYmhyneyDuqCPc+fLS8fRdnAAbXoHsPCZT5noMR2VQkVRYQUTXvieTz85AMDRDaf57csDtHjvWW44mWO7ZBhFSZXId1ymm89gEn3saNLFg7jrafRCTtiuK7R7bTBR8bk019Ywp/N8prV5H/PEFDL2XuT9fkv5YsSntGhjxykrDa3njyZSKaX51CGUXczFL8cAw26tWRfyJTsz1zP/l1n/e0Pq0UNdIe7Xr9+9fyyR1OoV44Z8hXjFihXMmzfvof4mPDwchUJBly5d6qiqxkVf+T1KdgC7d+9myJAhWFmJIxCA/vMTBIF3ui3EK8CDuZv/96QoOp2OzZs3086rE46e9ng0d7v7XGlBGRs+2EZMcAJfnliMo+eTcaPO/1LX2V36PZj1C3cwbO4oxkzpg0xmQMLNJL754jCxOVfZtn0NarWWT/61j5de7kGXrs3Y8Ol+wvdfxbmJA0VZRcxaO41quZwNP5zF08mAoN8O0WXU86QXVtGtiQ1Xj9/EYWgHijPLkbmaYe1swdfvjwbuTAyw47P9nNl2gXd+eINOQ9oCkBKVznfT12NmacpnRxc99nbqS30ee2G30rC2NqWJtwOff/4FPbuPom17L05dS2DX8Zt0dHAkJjIDmllAaQVCWg02LV1JKyujeZmS9OAEqvtYUmJsSYvmzcjOycNBYYC0iR3K3BqMnIxJU1XQpsaCzKg83JuYk3M4Ao9Ab7ISM7Hq2Qp1RjEVVgY4IKX4fAwezd1Ij8uk+3OduXLwOh7NXSnIKMKzhTtFOSWsurycwuxiPhy8jIGTejPhw9G1uk8elz7OnSe2nmPRvI/Yc3Y7UkMDinNKaN2rBbO6L8K2uQczv5pEfkUVn3x/GL+iKlLDU3Ga2IP0sgqamOTx6pDn+W7lMcy97Cg8H4lnx+akFlbh1tSMMHsdHVIg52wkU5ZNwK2pMys+P4S5uZwsQx1N5EZk5lXReXQg4wa1Y9uWS0SFZ+DgZElJSRU6B2M8fexZNnM4t8NSWDZuJR3eGMjF7EIWTRtCuxbudbJP/slDdbw5d+5cXdXxQIyMjPQ2d/2wYcMe6vU1NTW899577Nq1q44qanz0ld/DZgd3Pszs2bOHsWPH1kFFjVNd51eYVcTWj35j4OTetO3X6u7jf+YnkUjoOLQd0UFxnPrlAoNe7nP3NaE3Uzh7KppXX++Dnb0Fq1atwtjYmA6DAu9bj7WDFYNe7oO1oxVmVg8/CUBjVJfZpUZncOrXC7h62pN0OoyyZwOxd7Mj9FQElhoVTX3aExmRQbfuvrRs5c6Fc7Fo1FqKVDpcfV3QKlQYPd+as8oSjA7cxk5Q8euRXxjc5zkGDW1DbGoxAe5W5BZUUF2txVGtQJNSiS4ljwNrjjFq5jPIjeUMnNybrIRsjv50itxyBYmpRbhLdciN5HQc0pbqihp+mvcr7Qe2pveYbnWyL+pKfZ07BUEgOCgRW1sz9u29joVVM87FplJWUEzIziDcrUzIi43AoakHqhoJGFsidzdFdysFf19HBkzqzBFzFbsu/sxzEz/EKEOJTQd7FIY6LBIELOUyZGeTad3CnkGjW3PFyphk03Ice/iiRo6znSW6hGyMmrphKpcgScjAxd8dhZsDrkZykm4l06ZXC9oPDuTsmRjKpAbIiivZ/unvjH7nWboO79Agp3aur/zy0wv4eeluWnZvTtCBG/Tt1RdrRytsnKzxauGOTqej05C2xF1P5Nq+q0RcicfLwRKFrQ0uPU2o0emoLoznRlEarbx6YO3pSIWzIfa9fFGpDHFp4UiJkyEeVlLyjRWY27bhkryGJlHZWHk5UOotw9rAgDxjHbZFVqRWllOp0zBoSGuK1SqqdBrss6vRZldRYm/KwXMR9PT3oNuIjvi19kTnbIWTnQXndgUReTGGaSsmY2JWf8PpPVCXiYkTJwLg7e2Nj48PPj4+d//9tPjwww8f6vVXr15l7ty5ODg8+VefGrqHzQ7udJVYu3atOCNWPVJUKUmLzaQ4t/TuY6nRGbw2aRoatQaAcfNGotNqCTp4nbgbtzm26QyF2cUUF1WSmlqAQqFGq9WSn5/P9OnT7y6nuqKGY5vOkBiazNGNZ2jWzptpX04Sb6KqBZUllRRll+DfxZeM+GzObL9ExMVobp4KR1lQxrlz2wk+FUleah7WHpbk5JQSdDme0ioVPu28qahRUeNuSXJ+ERlxWZQUZ+Ps05bBU0fSrmszmphIcfW0w9jWgoqyaoTiclQ5ZSizS7l+NJSYqwkc33KO6vJqLB0syU0tIDe7lLTUQrJu51JdUYN7c1fUSjWp0elc+j2Y7KRcAPLSCji++Sw1VU/eWPqx1xK5uPcqD/IlcHxIEhd2X+FKWDLhkelcC00mNiaL777+lMTkXPIyiihKzseoQkl5fikauQRVqRJVqRKdRqAirwSZXMC5lQupJvk0nTgBM52MqoJqMBCokKhQKdRoZFoq80qwKdDgZWWNaRML8uVa8HOgTCOgszahuqgctVoLxVVUZxYjOFhSptIhdbOhtKQKy04+dHu+K349WlCtEVAr1YSejrgz090Xk/7yQ/DTorpCQXpMJnlphRRmFXEm6hg5qfmc3XEJrVaLVCpl3NyRaFQagq7Fkh6biYlSR1m1Cm2AI+XVSgoyknBpP5iw4NuoJAJlggadhx2lCjUSSzkllTWYqgwolupQ+dkRnpPHjeg0lBKBUqkanYmEIq0Crb2c9KJSTu65gpGtHMFRTnl1DdqiMjT5lVRUKLhyLYn8wgq8Atxx87DH282W8pwSLu+/RkpUOhqVpl733wNdId6xYwcAKSkpdVpMampqnS6/vpw/fx5PT0+aNWv2zy8W1Yqq8mpqKhXYu9o+9N/mpuZj42R1dyD4LVu2MGfOnFoZrk304Nz9XFl99bO7v2s1WrZ9sofkyDQSIjNw9bBFIpHwyicTWDBkOVq1lhvHwxg//zleXfYiAwe3RqFQsHPnTj58fwFlheWoFGosbM1JDE3m+xkb6D6qM5f3XUPQ6Rj2xiAAyosr0Kq1992kkx6XhUdzV/FD0T9o1bMFa69/gVarpSCjkF8/3kOz9t5kxGUjM5YhVFQR9stpVOVlBOk0dG7mxvWzt5n1Zl+GjOnMDRcZiWn5/PbBJCLsrfnsk730Hf4MI/q1Zv8Pp9g0Zwvt+rUi4kocPr2bczssDfMB7ZHqBDLPhrJ1yS6iLsbSdURHrhy8wfxfZtFvQo+79U1r8x5r3vuZdde+YOLC51kyegWmlqa8s+51fv/uCIfWHcfU0qTRXTX+J79+soe4a4l0Hd4BufHf3wOhUqjYtGA7CTeSkI7rjIFUSnF+Gc2N5JSn5tOlRsuUd56lqkbL0dMxuA4N4LqyDCdzI1ALaJJKcB7Sit8tSwha9jU55rb0tWzO9evp+LV0IfFaIfZWZmRSTZ57Fc3fbMnl7Tnk/3aRG4lZtOvmxZnyDDq0dSUoQ4G/ZRPyC0tQX43GqaMP1yRq2nrbcykxiy7D27HtWjzerZowfmI3rm89TWZKHoJO4Kd5v/DV6SUUZhXj4u14d6zqp0mTlh6sDv4cABMzIw4s2sm2j/cQdTkOD3837N1sMZAZMOGbSUzff5QeTVpz/WImfmN9OaZJxT06DhfDblieSiY7s5jyQS1xlVsSmZeJfytXrkZk0NXDgeDLqTTv7c6tpAy8bCy4YV2FSSE4FVsTnZRNR38PrkamMUxqzuFlO4lWVHBdUUyn8GLSojKQ9QrELKuG6Ju5bAhLJ2bfFVpNH8KV4hK6SQyJ/eMGyw4tAKA4twRbZ5t62X8P1Ie4X79+f/mmIJFIOHPmTJ0U9iDqc3D5y5cv07Nnz398XUFBARMmTODw4cOYmJj84+ufZrWZ39udPyQ5PJXD1dvvm2Thf2VXkFnES15v0f25jizd9wEnTpzg0KFDrFmz5rHqeRrU9fG3ZtYmDq49jtkQb6oV1ljHpyCVSijJLSWwfyu+OvkvDq47zpqZm5iz/k2enTaQOXPmMGDAAHbOPIKqRkV5cSVterdgxZml1FQpMDEz5sf3f+b3bw+z4swS2vZrxSt+s6gormRf4Za76z63K4jPJn7HWyun8Pzsh+9y09DVRXYb5m9j94qDfHp0IR0HB6JRaVAo1AzoN4fu3Xqwat0rHLgaxZKdp+hXJBC98yoey58jNK8AJ08rikrKSFm7lv6jZxDmK6Gf3IXkg0m0szci7MA1XFaM4GZpER1TjMnPKUMQBLyaOPDjptdQVCsxMTO+m/F/0qg1PPfsN9hbGJNzMIgW3fyIuRJPh0GB3DwVTuveLYi8GMtP4V/j3bpxDK34IPnpdDrUSvXdD/p/Z9n4lVzaG4xrMycKMovRabS4+rqQHpOJTWtTSiKr6TA4kJsnw2nV05/ooHhMlzxDRkk5ZiUCjg6W3FaV0USq5uLv2xi1+CMuFWUytMSB2OA0zJ73IKW4BLMaA2wsTUlTldHGzZnwzFx627kTFpJOy44ehKRk4eNrT1JBMXIbOWaGhhTlV9He1J7sY2m0dTEnfM9l2rw2hPD4PPyau3A7MRdLKxMMtDpKj1/Dt2NT4q/f5oXZw5i+ckod7PnaUV9tlzOnzjBg0ADe6bGIhBtJmFgYY25tRn5aAX6dmxEXnEjgS30JSyul0vE2qRVlDOo0muS0ApxLJcjkUoqLqvDo4sZF62IG5clI/P4CHl8P55KymAHHC0kJSsTa2RqdVEqZjyetAz35ZtUkqpVqTI1krPv+BPv2huA8zJvo7HyOLH4Fc7kRY577Dm9Hc5L3XqL12B5E5FQR0K8Z129n0imnnPgLsTj7OFJWUM7B0l/qbB/9pwdqEKelpQGwcOFCxo8fT48ePQgJCWHv3r1s2LChzov8O/XZIN60aROvvfbaP75u27ZttGzZknbt2tVpPU+C2swvP72A8uJKmrX1vu+5f8ou5mo8Lk2dsXG04qOPPmLBggWYmj5e39KpLd7FM8Cdpb8//M18jUVt5KdRa3it5Ryad2rKwu2zCT8fzSdjv+HtVa/SZVh7Vr21gX0nDrD2wI8425uz8IPfsDM1JPdiJB2HtOXakZv4d/Ej7loCgYMC2H14F2ejT6CqUbFo+OdY2JqTn17IoMl9mPblJODOlM5r39nC9eO3WHV5+f83HDQ0afnvO7+1Gi0RF2No2b35/7y61lg9bnaFhRXMfHMLAwe35vU3+3Fm+yXWvruZlxa/wPPvDrt7AeWNVzeQmHgJxywLWo9ox2W1kgAvJ5IvpNCkgz0hxgoWD+tPgKcT0wa+jVQCVS/2xtfYjKIVFwgY1o2Y1EL8mtsTpqqheY2KwqAUzLu1QmIoIVeuJbCZK7EXk+nW3Zegy/H08LEh+LcgAvsEEHf9NuvDVpCWWcIn/9pHU0dz5n78PGnRmfgEepEUlop3a08y4rJp3ev+KaQbqsfN79K+a3z35no+3PYOfh18+GrKWpIj0jC3NkVqJKPMyQEPCzlnDv/GlFde5ZWPx7F11QnOX0/h7dd606J/C97dfhhBraVMqcDdzppz+3cxoOezpGfU4NPTjbCcHDpgT2JlGc4GpqiUGnK81HiaWJEZX4R/M2dCi7PpaupKWFoOAe6OJOYVYdPEgnK1EmmBFmdrc9JzSwk0syP6dg5tbS0IL6mkra09KQm5mHdwoKimBss/YnBxtuLDbe+QmZBDs3bemFk23PsDavO9T1GtZFrr92jTJ4B5m98m5GQ4n7+0ip6jO7N913aeH/EC4eejsevVhhqlBqVGh5OZjIyzYfj2aUVshQb/1u78vn8jg6ZPJzWlEC8DU6ryq9DqBGydLElWVeLe2ZEQRS79U01IiC/lxff6MdCvCR/8epBMRRVOv8RiHeBBiqsV7bo04WxuGu8P6cXQFs1Y/e0Jrt1MwlWmoTIsA7mxDLmzLfkmZrTxtmXmolGkpBXj4+fEml0XuXwlDn9BTl5RNUs/eR5nL3umt5tHj1GdeXvV1FpK4X4PPcrEf95Y17dvX86fP18XdT2QhjZ188GDB+nRowf29k/PFKKPo6FM/Qt3bibZsmULU6fWzsH2x7oT2LpY03P0kzvCSG3kJwgCB9ccx6mJA91GdKQop4SjG07T78WeuPu6EHU5lucnvcgrry7BuuROv9Bn3h5K0tkIIi7E4NzEgUEv9+HAhqOERd9i9KhRTPrXWIxNjTiw5hjnfwvCwsaCcfNG3tPgib2WSNjZKEbNGsrOz/dTXV7DzNX//IH3SfG42SmVarb9fJnoyExeGNcZL2dLfv14DxUllbQa04OUjBLsKytIiMlkf9RuJg6dyTOj2xNfrSDA05Ez52JIrqrA2EqOhUROVWUa+bnVGHu6IHU0Q1mtwqaghkJHOS5lRlRWqbHOK0MTk0XnZ9pjG+jD8WMRGJjJUFWpMNPpqI5OxdTVluqMAkzNjampVGDk54F9Mxfe+2AYu38NImTHBTqO6kiYUsmYwe3o3KpxXBH+b4+bX05KHqd/vcjQqf0JPR3BiS3n6DS0LdZOVkgNDchXCoRsP8/R+D2MfH4+3fv4c+VyArLScowqq3F4uQs3M3LRGAvIDAyIvXSWFj37UCnVYasyokxQYuQgp1JQYptvhEahQeEqIDGSIs3QYak1hJQqNO1MKbBV4FBgSk2FGpmNjGoDDfaxaiQGUnI6CFgo5VRWabBTyKmsVmEjyKkpV2JuaEiVVIdJpRZZUTVaqZQOHb25feImQ6b0o9uIjnWw52tHbb73rXzjR3KS89FqNAx+pR8dB7fh5yW7SYvJ4NjtfTzj+wLtB7TGrUcAx1cfQWpoiKq8GhMLE6oc7TAylREReZmW/QZSbKJFbiWjRtDwgmNTNNVqgjafRurrQKGvDXJ7GVWVKhyqDBHMDels4cCN8FSUTnJkOVUYlGuRuJlT4ARW5WpMz6fRdXAgV4+Goe7VlAqJBrt0JYJOQJuYidzdHo2dDR0HNyeoLIfuzu5cDr5NC29nAt0cKSqoYNQLndBqNBz4/hjN2nvTrn/rWkrhfg81DrG5uTmff/45p06d4uOPP8bFxaWu6mp0EhIS2LhxI7a2D9+HVaR/mzZtorCwsNaWN3LGkCe6MVxbJBIJo2Y9c/fNy87Fhsn/Gou7751zS6ueLbC0syA/o4jk8FQEnQ65sRy5pxMSQwN6vdAVN18XDofuw8zAHEMjQ0JPhXN881menTYQW2drWvX0x6ulO3u+/oOwc1HsXnHwTmPJRE5YWDrxMVkoa1T31ZafXsDuFQcpzi2p133SGBgZyRjxXAeMjWWo1VoMDKUYGEopKywn4mIsuakFJIYmo9PeGf83oKc/rXq3QOIgR2IjR+ksR6fWYlQukJGVzb6ff0Lu6ozGVI68XIeg0oGhARqpgMrGAI2gQyIIaCRSZJ6OGFVUIdQokRZXIVFqaNXaA5mBBAkgkxvSumcLEASQSFApNYCAvKQcdVUNYbfSKCys4OyZaPLzHn2M8sbMxduJyf8ai4O7HWqlBq1Gi8xIjpGJHCNTI9QeRmgkAsgMqdZouZKQRk21EklZJYWF5UTdzkUj1aE1EMiMDacgORGJuQy1VIfaTkAj6DBQgwYdgoGASi6gMhZABzqtgM5YglarQyIFjVSHYHnnbwRD0AoCglZAohHQabQYFlSiU2vQWEhRo0NnIkEr6BAkoNPokKgFBIkEjVZHZFAcBRlFXNx79e6Nk/EhSexbdQRljfIf9krjpFaq8Wntgbm1GRqVBns3OwZO7o2ppSkWdhZY2poz4q0hDBrSBhcnKzQyHYIBBHRvjtzUiEtXdmNgaAxSCcbGcvybOaMxkqB1N0ZwNkbD/x97Oh1SnQS1VIfC2YCimhpC0rLQqLSgEVA6maA2lyEYSFCjQyIVqKys4Vp0KtUaJUoHAyycrPBq5YHKzgS1qQFqMyk1goYrKWkUlFYSEpKMslqFqRI6d2nK82M7I5VKkMlljJ07sk4bw/CQDeJdu3ZhaWnJ/v37sbKyeuTZoxqj/fv3/8/njxw5wrp168TpYRsQQRCICU7gt127//F18fHxvP/++/VUmehP2Um55Gf89QcRhVJNTFIuJ4/+gSQ+lfyMIrKNzTiw7wYHjkbg0q0FnZ5px8m9Z9DlGODj3pSD60+z8/vjbFu2F0OZAS8vHc/Qqf3IiMtm2/K9nPzlAtuW7WX3ioP8+skedv54hnQDY9754Y371h99JYHty3/n9q3UOt4LjZOjkyWfrZhA3/4BRF6MJWj/ddyauRB/5AaSzDzSsgrJtbdm6KCZPDu8LZfOhrPu+GU2Xw7hSFICluZG3L6RTkZWFB59nkNbKiBNU1JwJhPHNCXpq87jXKHlWnYmtqbG3ESD0NuPHYdD2PLjYdQRt0k/eh3r/EJGv9Sd5z9+kahiJR2nDmLk+8MwbeVC0tlbvDCxC8lpBexacwRDcxPiUktxKlJy+XAkMdGZ+t6Nejf8zUE4ezuy++uDbFu2l23fH2XrtXCMJrTHP+AlTDzMuaDMwby6jJjLcUi6Nyc1rYg8xyoKTauICb9J+8mTuFaegZWtEVer0rEzMiEmNAfrCiNiigoptlVSWFVNcnYJGje4IsmmZrAFl6tycMCMq5o0TM1lxKcWoFZoiPJRkqCqwOamjrxPg3DKVHI9JxMHE2OuF2Rhq5aSfiUVoaiSZF01GUEREJNCwuEbmNuYEbT/OuHno1Eo1BzYcp7ty3+nNP/J+vBTnFtCelwW83+exVvfvsonB+Yz/M07Nwq37N6cl5eM5fyVs0xeMg6psZy01ALGffAc0W6maHt6M/qNAXTo6EZJkYJ2TQNJuJzK8ABfXu3YHoMqHT8du8rmEyFUdXQnskaFuZWca4VZ2FmYcKUmHVMbOVeNiilpKiORcmItKyhqJiW8rAhLqZwgypCNaU+QJWgnteG6rICxPVvzwpTuRDdRkTO6Kde8DNA2NyOoJhfXGiMSz6dih4zDe29w6+a/B3FIi82kJK/0nu1XKdXEBCc80CgqD+qhukyUlZVx4sQJampq7vYRe/nll2utmIdVn10mXn/9dTZu3PiXz23fvp0xY8bobYzkxqqu87t25CaLR3yBMrCIi7f+egIZtVrNrl27mDx5cq2v/0lXG/k9bzcFM2szfk1ae99zX246xf6zEZSd+gWPoubsLdiEibkphjIpt85FMX/gJ3R/qR0D3+hF+J5EDq45RrNXBnE7uYCde2ciNzBgrONUWvb0Z+X5T9BqtRgYGLB+3i/s/eYQ/l2aEXftNjvSf8TW2RqpgfS+m4f//BuNWoOh7MkZdeRxstNqtPfsqz/3zZ/76sO5O7l5IxlJb0cUCjVJ2zbzbM+XSf39Ek1n9OaitYYu3u5cS8nEMSeLAqkVRz59G0dLcwZNW4O7ly2RlBDY0pmrynSGZbsTfzwVh0k+hBTl4u5iRXZFBTEzZpEanc70tvPoO6EHi3bM5ufNF/h162XcxjblRl4OrVXWZOaWYiCR4GQko3TjeQK6+xFzJYGXFr/AlE8m3N0ujUaLYSMZmaBW+u//1/a+1moOOcl5aDq3wrupI+s3vc6zz7yAqqo9be2NCT8QTMDUwUQn5GPw4p2JOCqvXqd1n4GkFZfRxtuFm/k59LJz52ZUJi3bu3IjJxt/M1tSC0vQORlgb25MhqyAQCs3bmUW0Mvck5uxWfQ0dyEiMgv3Vg4k5Raj7QLmghFV1xQE+DgSXphPP7kz8aeTadPWk4iwdJopKkgJTUY5rx8ylQ7Jdxdo2t6b2zeTGfn2UGatfo11359k394bfPnNi3To1HCGia2N/GZ0mk9SWCqHKn9FbnTvfQ6/fXWAjR9uJ9fpNs55zfCeNJDU9CJMw+MxszIlP72QJv38uZ6TzKCWHQn//QreS4ZzraiY5m4OJGYX0iRJh1wiUHL0Gu5jAznfzYR+MneigzJo38Kd0NhMmns7kpCaj4mXOah1GFwrxjPQhRDK6GHjRMK5FALbeRF+Kw3PXl7EpuUh62pJlVKNLl1FcysbCs9n0qJ7E25k5NI+wIPQmAw6WtsTF5nJrzvfwsbWjOcsX8a3vTdrrn1x93zz85Lf2LZsL8v++JCuwzvURiwP1yDu27cvfn5+uLn9e9anJUuW1Eohj6Ih9CG+fPkymzdvZvPmzXW6/idRfeR3OyyFOR+9w6FDh/7y+Y8++ohWrVoxfvz4Oln/k6w28svPKMRQZnDfsDoH1x7npw+34f/6YLZv+4ITBw/QppsvE174Hlc7M1L2XqJpey9+Ob+Bye8s4maRmtbedtyqKCcwvIji+CwMurXBQiYhp6CSgc8GMmvOUADUKjXpsVm4+DhRmFWMp78br/jNwtbZmm8vLruvxq0f7WLfqiOsDv4Mr4D6my6+Lj1OdhO9puPu58pXp/5FxMUYFj7zKW+seJmRM4aw4YNfObTxDDazB5CuqOLX915k8tgxWJkMwaeJNSGVFbRt78mtsHTcm8g4/uuPDHtrIdklFViYGCEXJJRmVODt50hYbi5t27pxPSGDzl5uXFHk0E7iSGpWCeaOxpiqQfLVJdy6NCPCxoR+LnYErz5OwKxnCCksZtHMoXRq5sH7Q5ahVWn59uxSyjOLcPRyICsxB+9WnneH5vrmy8NcvpTAlm1vYm3d8Memftxj78ypKL758ggffTyabj38ACjJL2POW1uRGskoORtKy36t2XB6Pa8+N5vw7edoPqM/wToFrZzsiS8vpSQrBGVpKS4tB+LlZkt4dT7tm7pypSKVnhZeXCnOoLOhGwlZhThamFEjaCh2Lcfbyob4kgI6mHhx62Y2XTs6E6JNoktiMxISC3BysKRcXYNSqcXV3or4shLauziSdCyZgH4+BBXk0Lq5C2ElOXQ8mENxWRVrrn1OelwWs3eeoL3SkKSfL/PWt68y+NV+rHx3C1d3XmJ92Nc4N3Gs7SgeSW2cO8uLKpjbfymCTmBD5Mq7j189FMKycd8Q2LcVm0/+wI5fdtJpREc+eG8HFXmlKK9F49TGk71Jf9B/8Fiya4zp6GrHjdwiWrVwJ/52Lq7O1pQWVlLuDJ5aGaVXC3Eb0YTz5DBI6knYjXT8+roTUpmHr9yG7OoKJEhww5TiyFJ8WzpzMzObLt7uXC7IoJu1G2E5eXg0syNNUYbnllgsHW3IN7agmb8zIXn5BHby5HJBBh8M6MOoTi1Z8uJ3pF5PwMzKFANTOYX2DrT2cyJs62leXjqO52cPIyUq/S9vpH9UD9Ug7t+/f61O1fy46rNBvHz5chYvXnzf40uWLOH999+v8/U/ieoqv/LiCj598Tu6j+zEc28PZcarMzFOseWtlVPwbf/vqwQ1NTUsW7aMzz777H8sTfR3Hie/Xz/eQ0JoEv/a8z4qhZpPJ3xLuwFtGPv+CODO2NA/rz1FSlENReXX2LHzR76fvp7MzGJUFTVYWJoQnxILplocpZ5Y92lFSaUaC0czarLKGdnXDwNnOw78fgNzmZTK0AS6DG1L2Nko2k3oxa3YXFraGJF0I5FFu+YQciIcUwtjOg29f3SYtJgMIi7EMOiVvhibPhnfAj1Odud2BWFlb0H7gW2oLK1i26pjRGaW0bW5I8G7g7Bo7U2ZTIbgYYajjQXHz2yjU6dh6A5EYdy3GUlN5bgam3Hp+O+06tQLlYUF1jo5Ko0WqgXM5DIKTJQ4GphQnF+Nm4ExVQml6IZaUiCpwUppSo2gQRAErIo1VKrAztyc/JpKvDO0lBdrsHGxRFOpxrGHG9HxGVinKmnm50LW2Vv4d2pG5OVYxr4/8m7f9bjYbFKS8hk8tA0Ghg2/29vj5JcYmsz3szZh0s2PEnMZC6cNwdvNjh/XnCY6OgOiUzAxN6LCyYGo5NN4d3kWVyUUV6oxbW5LcXk15uZGXDz0M75zx2Omk6PM1WLpYEyuWSmOJhbkKypommyOMqoSZR9LygU1Ht5FYCAnOdMEewMTqhIV2DiYkaOtwsFDTn5lDXYqU8pRYR9vgAESyK7GzNyAikuxmPX347a3DEdLM/K0VTjqTKkuqORVZx96je7Cxx/9jspcSqFSSQsjOa9NHYCjpwPJEWnEXE1g8JS+yI1kdZTIw6mt975rR26y8/P9+HduxvSVU0iLzeSb13/Awtqcwuwi4qrD6OjSgwU73iU5vZgd264g12jo2duVnYcPYZlkgWVvf3INpdg4m5NfU42ltQnlGiVOtzVoTaVkdNbhUGVEvlaFo4k5RbmVONiYkmVUhb2hGWUKBRbVMtAJGGSpsTQyprSyBnsLEyqSSrHytKQ8owJpe2uKBCXmggxJYQ2SSh0WFqaUVCuwNzOhIrUMyyaWqHJqaOtgQujxWxibGqFWqtFqdZg62VCRW4KnrwsvLx2LW7Pav4ftoY78wMBADh06RFpaGunp6aSnp9d6QQ3Viy++eN9jW7duZenSpWJjWE/iQ5LY+tEuyosr7nncwECKnYsN5v9/pWfksFHYOlsjN/n310rl5eUcPXpUbAzrQejpCOJu3MbK3hKJRMKB1cfuTqDxp5qKGlQVNTg5WTJ//kxkMgOsHa0wMDbCyNKUdEkR1s4OvPLaK3j4OtO9Y1MsrEzQGYCRoZSSvDIKNEpkZnI0xlJM7MyxtLXA2MyIqPPRmJnIsbQzx+b/u0r0GduNwL4t2b78d0JOhgOQmZjDpoU7MDE3ZsRbQ56YxvDjKsktpSC7hF+3XuJ2cgGDJ/XCwsaM8JRcDOzMUDqb4+lhR2AzV2ytTOk0YDBKSymGjmYoLCSYyuXEXblIQP8h2BraYmpgiAQwlEmRy6TIcsswVQlodWAmk4FCh7GZHBRgbCgDQ5BLDJBLDdDYmmBkYYxGImBibITCxRxjcyM0CGi1WjJu52NoaYzEzhQrGzMs7S2JS8xDZ2jIufNxZGUWA1CZUUBxeBJarVa/O7cOHFx7nFO/XLj7u9xEjqW1Gbq8MgwUag6tOUZVWRVW1qYYmMowsDZB7WqJzNqEZm16YGxuRLWzGXJLI1RGOtwcLFFmR9Fz+MsYY4ggk2BkJENSo8NMIkenFDCVylEbCsgs5ejkIDeRIhgaQpUhJsjRasHYWI5KrsPESIZaK2BkZIjGAgyNDBHMpUjkUmSmMnQ6CcZWpghqMDeQIanSYa41xKhIi5GxMQVWxhz4/QZqtZZOLbxwsbNAWVZNYnwumzecw9LVluFvDmowjeHaVJJfjoHMgLSsUs4cC+fwDyeRSqVoNBpMrM1p7tsNncyQA7+HUBSTjqFCSam6jNuZWfR06ICpiRxqtBibyaix0mFkIkMwEDARpBgKUqQGUkwkhigddBibGqJT6TA1lqMyFrDQGmKUp8FEYoihRIpxJZgKBkgVWkxkhqgMBYwt5CiNBYyNDDEs12JeqcE4uxK5kRwjEyO0BhLMzY3pFuiNnZ05Qo0OqURC/O18ZEYyJFIJfh2b0qytN4Gdm+LiaY9aoaKmom5mlnyoBnFYWBgrV65kypQpvPLKK0yZMqVOimqI5syZc8/vv/32G0lJSeIsVnoUdy2R41vOcnnf9Xs61ptamjLo5T607d8KgNVbvmPQuyMxtTTl5qlwBEFg9uzZODk56av0p1ZxRTUHT94kJTqdns93JikslZNbz+PoZc+gl3tz4/gtCnNLOPFHKJGHrtHW1ZyFiz4kMjydPuO7UygxwLCNB6eijuHWpTXdx3al18ej6T6gJZbWUqJ05aiqyzixM4jDJ8NRqKpIKinFZGgb+k3py5BX+5N+JZYxz7Zi7DvPMm/z2+SnFXD7Vgo1lQpO/HyOY5vOUFpQRlp0Bid/PseZHZfQap68htKjSI5I49D6k1w+FMKJY+Ec/fUilmZyPNq5EocWh0ldidVW0aWXHwP7tmBwvwDCt20hubCQqnEtCDdTo4lOIf7yNdzzTSi4motRtYSSgmrUWh2qGiXFh6KxzK4iOaMAC7mMuPwilF3MSEovwbRARmZuOcoiFcpyNdl5ZZiXCaQmF2BmJiO6tBidvznxZSVILORURhdjIJVSpKukX29fOo3oSEqVBnk3f0KTCzh1IpKCrGKObTjN0Y2nCT8fpe9dXOvO7rjEkQ2nyLqdy80byVi52DDsjUGkn4nENiKLK79eJOhgCC9O6o7SRkZpoAdxLiYI9nKOHdqEqbkRkRVFaNpYEC4tJinhJC1buVJirUGdqiMzvwIsICWlGCulCbGZRViozQg1L6a0n5yUshLUWiVJhQbExuqwNlCRpMzDwMOQSEU+5tYy4ivzMJHKSKkuRCdoyHavIMOolPKmcuKpRjcsgLSscuwEAzLCC3DIkVIUnIcsT8ehfcEc33MNqVRC317N6WwoI3b3VUKvJHDyeCR5OWX6jqBWleSVEno6AkEQuLjnClqdQJ7chKM/X+Dy/ms0aelBWnQG3v3acOTGPmQeDpy9EMeR3VcpTEzl8PGNZCRVceXgDUzsrUhJK8XEXkaEtgBjI0OykoqwiFWQp6wi2ViJrEBKalU+dgoDbqcVYmotJz67CPs8Q0rP5WOcJ5BbUUF1fAmCSkPW7XwstBKia4pQt7MgRFaAsU5FYXAONlfyqTwRiyKhHMrVpOcVY+lkSq9e/nh2cCe+tBTPpg5kqsDWzY78tAL6TujB7B/fYOLC52k/sA0xwQnEXb9dJ/v2obpMNDT66kMsCAIffPABn332GTLZk/eps77URn5fvbqGM79eZEPUt3j63+nbnpuazyvNZtJ7XHcW7ZhNp469sTIdQnsrA8KOhvLJqXmcCz7Ne3Peu28Wp8rSqrtXlkX/28Pk9+d+/XT3GfYERdAlOI+M8DRMLU0xszJhc+wqbp4MZ9Hwz2n7Ul9upZXS1sqQiKM3SXCLx7/ZBKziUzCUGTDqq/6UZVby29yj+M3uzzlTBYOTtUTvuobTshHEFJTglylFW6Ok8nQITfq3IszOhJH9WrPgtcH3Zfyq/ztUFFeyN38zyRFpTG83j2enDWT2j2/wy9Ld/PrJHpYc+ICeIzsBoKxRIpFKG/UVpwfN7j/3lVaj5eVmM9FpdezMWM+tCzHM77+UFlP7cq26hvlTBzJ6QCALP9jFzRspaJ91QFGtJOHjzxg+ZBrXOlrR3c6NY5s2MujFiYTmltCyoxuh2dkceX0yjmbmPDv9B1ybWXPdsZDuZh5EXc8msI8rFytS6V3tRWR8Li5NrMkpL6fGTqCpxBL1nhy8hnpyyiaXfuZeXIvNpKeZC1HR2WxeMgEfL0desJuCnbsdOYk5+L7QhRBD6O7pSkRQMu0tDQg7FkqT1p5kxmVxoOyXBp/twxx7KoWKUdav4NXZl9sSIwYOasX8xc+xaeEOdn2xn5bdmxN7LZEfIlfi4uXAkOnr8LCzoiA0h8TMvTh1eYm2Hb24mpKBX3t7Tm/bSse3XiS9ugx5rhwXKwuSFSW093QipDqV7pbNuJKUSWd/R0KKMwjMcCM9txTpgCocjI0RZOl4aJtzJriGbk08uJqaQYcWrlzLy6C1tyVJlYV42Ggw0pmQECKnlZMjYTm5dG3hTFB1Ct0Nm3E9KYuuCgeSUwpwPRODgZGcYi93mjmZk7D7ImPeG8GbX+vvhv9/8jjvfV9PXcepX87f856XGp3Bm23n0m1ER5bu+/dkUAGebXDP8qfpy4NISS+ixKII1GVcOfQj106Gs+TZzwh8ezCXi8tp16UJV5PS6VJtSVpSPlWDbJAZGpCjq6aNhQNpQTm07OvJueoMurl5cDUrg96ljiQmFSDpZI5OJ1BZVEMrM1sKjqbTvJ8Pl4pz6V+oI2pXMM2nDiExqQDT1nbUqNRUVqvw8rEnTFtIN3s3oq5nMHN8LyaN7ExltRITI0OmtXqP4rxS9uRtJishmzfavM+AyX2Yv/XtWs3jTw/UIJ44cSI7duzA29v77hVRQRCQSCQkJyfXSWEPoj4bxGfOnGHAgAFotVp27NghjkpQC2ojP61GS3FuKQ7udvc8XpBZhI2TFYYyQ86cOUPLgA44OFoQERJFcWUhu98/RnlRBTsz1t/9mysHb7D0+RW8+8M0hr0x6LG27WnwoPld3n+NT8Z8w+z1bzJkaj+Wvrqa0D3XWHX1M2ydrJjS/B28AtxJupVKwLMdCS9S0L6jN6EhKfh1cefSrWt0LrWk8HYO+apsfJv5UZBQQkBXPyIuRtPm7aGEh+fSxt+Z6OQCzIa7kVlcjttNBY42Zny/YSpnIpNZ+OsxnpGac3PNKZbsm0f3/2/gVpVXo9VosbS1AO7c6GfnaoOBgQGCILDykwOcuhDHug1T8WnqxCTvGZiYG99zI0tj8yDZ7fvuCOvn/synRxfRcXAgs3stJjUyg1+S11Cj0vHqpB/xbeHIjeJiJo3szIzxvfh66xn+OBeBcxt7UqrKkFoaIo2MxybFEe+unpy8cIQR/UYScSmJdxY8y9BBbXhl316iCwuw0hphKhhSWFiBbzMHrldn0dPBi4sFqfRt7sKt0mT8SpqRUFSEtZExhjoJRieK8QpwIsi4hJ52LiQeuI33S35cKs6i/+VK0i/EYTuiO0qFkjQPKb6WliSnFNM2wIOQ6Axmje3FhBEdWPr819w6HcHqa5/j3cqzntN4eA977izJK8XEwoQb56JZNupLuo3owJUDN2g3oBURF2Jwf2sQyZXVuFRIEYwMSHbQ0NzFjkthlxnQZwBXMtLpZOvAxTNnGWLoQ/qleJSL+qDTCRShwNfOjvjUfDq0c+ZidTI9bJtyJS+NDnIP4jIKcGgqp9pAicawkmaWdsSU5tHJqgmhlbcJlLfkQlY63SRexKTl4eRmSVlNFVVO1XhjT+aVcjobm5O04gS+y54lKLeEFr1cCC3KoVWRLQX5pchLBNzMTcjfcpFeozrx4bZ3GvQQqI87yktRTgmOHncmAKuprGG86xt4+LuRFJbC6HeG8ebXL7N5fzDfrNnK2o/epls3f8Y9v4RqpRatlxveXvaElxQyzK8JS6ePIKegHCd7Cz55fS03fgvGeXQviiqrqXCU4eRqRWJ5CW09XbiZnEWnVp5cj06n9VAHblWm45vjTlZ5BRaZAo525mRllRJoZkb8+lMEvPcs17OKaNnWiRsVRXQ5V0B2bCayHoGYO5oS5aqkpbsTt9KzGWDiSsyJ27R7pgVXY9Npn1ZMSmgKxr3bYmVvyYat07h8KpKPP/2Dl6f0ZvKUXrWei3iF+AFt27aNSZMm8dVXX2FjY8O0adPqdH1Pg4fJrzi3hI9Gfknf8T3u3nj1oP7MTqfT8dxzz9Hetju5EUUYygxxbepEbmoBXYd1YNQ7z7B9+e9cOxLKnJ/epFUP/7vLWDL6K4zNjFiw7d1H2tYn0YPmV1VezZWDN+jybHss7SxIj8siLTqD7qM6YWBgQPDhm1g7WhJ1OY4jG04TOKIzz701iKTEPLz8nJg3ZzkO5gFo1OWcufwz3ce8i7OZKTnVCpxcjYlyUdAhzYLC2BJMOtuRRQ0O5TKo0mBwMpKOA9swadl4ftkfzJWDkTh7WJKJhq5aKQlBcXx+bBFG5sbM+mwP/j7OzJ7U927t53YF8euyvTj5u1GeUchnRxeSHJGGTG5Iq56NZ6rf//Yg2ZXkl3HzZDg9RnfGxMyY+Bu3WT/3F6wcLFn823v8sesqf3y2lxYjOvDagtGE/hHCzq8OYDK5C7laFbY6OWqNloTQi3ToNIBrURfQFebgHTgMZz9rbptW0sXbgyvZ6TRxU6PSqqgKtsTSwpgsgypcHSxIUZbS1MGCdHUBbgb2ZBZXY2dmTIVagaAxwE4rJ09XjYeBNQUJ5fjrzMhLLMShmS3FicXY5hdQWVqNYmAAFtamvDWhJ6HnYzn95UFcBwSSW6aic5em3LiayPODAnhmcq97GlIpUel8Mfl7xs97jv4Ta/8N+FE96nufskbJ5X3Xadq2CbdvpdA00IvU6EycO3qzZd42KnJL0bXwwdjBlFxBSUlWGFYugXjZWXJw03f4jxuMqpk7zgo5JUYgAcxNjMg3qMTd0JLMwnKa+JiRpi3AXedEZmElzjIzShQKjFQSTEwNyLOooomtMTWyTKx17kRlKGhibkNmcQWeboYUCxVIVCYYSqRoIsDRwpycojI8TY2IM1HQ1MyGjNIyXEzNKSmsxlpliBQokWlwLVNRdTiKsXOfY+ir/eps/z+ux2m77Ft1hNPbLvLJwfnYu9ry4ZDlCIKOqZ9N5FZwEievJtOtpQtXd1ygwCIfN5UrH/w6gxenTOKLL9aikRuxa+MlbF0tKcwso5u/IyG/XWbWmtexdrTk24OXyFCrcD51G8zkpA/2xMHSjAxlBc2kVuQnlNCkRk3h5ThsRrajPFtJ2UBzUAvIY9TYm5lQFl9E6yZ2PP9qL34/Ek5USCoWrewozqvA8WYqakFCaVcfHN2smDamO2cib3PyejwttGbk3S7GJsCevKJynELTUKq0VPbyo22AO2+P78Xli/EEtvXE3qH223wP9BEqPz+fDz74gNWrVxMfH09AQAD+/v4cPXq01gtqqHbv3o1KpSI/P5/XX39d3+U8dYxMjQjo5odHc1fS47JY9dZPpMXeO6j+3pWH2PXlAQDKSqtZ/d0JTu2/wReLV5ASlU54eDgtPdqgytNi4emIqbs9vh2aEtDVD88WbphZmjJwUm9a92qBtaPVPctu3qkpJXmlHN/ccEZZaSzMLE0ZNLkPmQnZfP/2RqKD4ogOimP/nhv8uuUi0VfiSY3OpN+LPfHr4kdxtYrbN5OJORCMlYGU/NxIrFytSCuKYWDHsTjI5WhM5TjYWaJVSnHXGqNOycfOxgxJhRZ7mQlyiRTT8mosbM2IL6vkzOkIik5G4WBjilKQYmtqQmpeOb4dfZAZGbJz2xWMBAnNPOwpzC7m+7c3EnM1HqcmDrg1dUJTWoVve2/kJnLisiuIz6745w1v5GwcrRg4qTcR56NZN3sL7s1d6TaiI2XGhuw9GkLaxSjsnK2ojM3k1qGbXP49GOumjqilAg6GxpgoBIzMZGQn3kQjEVDkpxM4dCyOFiZoY3KwkxmRUFyIg9wUoUaOTGqCmZ0RWpkEB3Mz1AodrjJzhAwdzoZWKNU6HExNkUgk2OhMsNEZI9FKcZSbo9JpcTIzRp2ai721KZSosfS0QuFni6ylA0YORhimF5N0Kobs5GLsfZ0pNwWvZg408XbA0caU5Fsp7DwSwub9wWxcf46TxyOwsDEjoKsfTg1kuK7HVVlaTfSVeBJDk4kOiiMxpZCo3Epun4tBUqXE1ESOBVoMVQJuVhaUpYbT3N2BivQ8HF3dsPf1x8HMAsHcCGupEdYGRhhowcnAFFmGClcjM6rVWpyMLVEZqHA0MQU12EqMscAIJIbYy8xRFUqxFuworwRnMwsUghpnQ1MM8iVYGpphLJMhl0mxcTCmxkqDnbMZpV4SnGwsqJZocJaZYpSnwaFGh1lRFTJDQ2xMTah2MsPK14krB6+TcDMJgGtHQ1k3ewtV5dV63vu1w83XBWsHS377Yj8qhQr/zs1AIiHhRhKZUenYmsmJvxqHrZcD0VkRNO3QlCs3rjB82GQy0ypIPnQTG0MJQmoBduZyElIKsWziwO+XowkpK6HaRo6VpQkSLxvkah1ONQLyfDVuUlPIU+BoY061mQF2Xk5ISwRMXc2QSwwxVxtibWGCSi5g42pJvokBcTklFFVVY+1iiUomYOFqgTrAHmN7S2xMjFBKtcRk5pGVW4aTlTlVEh2+/s706+yLvbMV2qZ2mBrLsNFo8fV0oLJCQXRkJrcux7PqrZ/ITMiu1X37QA3il156CT8/P6qrq+nTpw8HDx7k6tWrfP7557VWyI4dOwgICMDX15c1a9bU2nJri1ar5Y8//uDrr78Wb6TTAzNLU95eNZWuwztQlF1M+PloTm+7SE3Vv+82jbueSMzVeLRaLef3XOHW5ViSY7OoKK7kyqWrFBdVYkZTyksqsWrTFFlTd8bNHclb306h1wtdAfBp48XsH9+4O3VwSmQawYdv4t/Fl7KCChJD7+8iVJBZxPnfglCr1PWzMxqp3JR8Ii/FkBSWSsjJcIIO3uDKkVCuHb3JyYh4CjUqPFp5kpxfzolbCYQEJ3D2VBS2dubklMdi1747gkqCtKiSguJKTGsEymNLcYnUURiejs7eEHVyBV6CGdUphaiuJqGztSLbWMbZq/HEhiVjZCChLLucADs7ijGk6eAOJIamcu3oLWRJRbSwNqeiuJLISzHkJOfj3MQBKwdLSgvLGD9/FCZmxkRHZRAXm6Xv3VkvCjKLOLcziBuX4wi6GIf/oLYUOVhz+WI84aEpSHydSUkr4NyuIFJu5yFp04TC4hrMs1WUVlZRaK5GEHSEXTrFwN6TKSyrRmYoUBGUjnWBhILiSuTFUrLSIL9ARqWzilzDCgxMIaeiAqsyOcUJlciURmTXVCA1gKKaKrSlWijUUZ5Vg1GlAbnllchkOvJi0sFBTkFWITp1JRk+5hT2cKPEXEPx7Uwu7QsmJiodgzbeZJrAwLHtaRvggmleEXFBscQk5RKRkEVkeDq3E/Owd7Pj3R/eoGX35vqO4pHEBCdw62wkcGcClXM7L3PrbCQpkWlEXY4jOSGX0JBkLp6PpbhKhVIup6aqkqKKKoyNDDA3NcI6J5fkiEi6tR9AUUI+1qElFFdXodaqMQgrpComH6NKKRVxZZjIJWRWlWOkNCZPV4KRkUBpUTWU6qiUqCgwqMJCpqawsIrqYhvSSqsxNZeQqyzHRGlA0W0lNVUGVOdrKC+pQWOnJVNahoGVlCxKMbI0IK+4DLsMLRVpZRhmllGdkEt5RQ1GRRoKa2rQdvIiNTmP079eJC0mk3M7LxNxIQZF1ZMxdXOXZ9vj17Epcddvo1ZpmLjoeRSVChJDU0i8Fk8zS0OKMgqhkw8KCZSZlODl5E1Zko6rR28SciwUUrIoCr2NUX4x+cU1GLbxIb6wgricfPJTczFMLia/mR2lgoBpfAWVsaXYZUFZQinGcgPSTSQIHZuSp1BQ5mmIIr0KXXQV0sxi8rMLkaqqSMnI59TVOOKLilE1NSFLXYXaDNJbWVPqYYu0RE1OeQVxOQXEp+Rgc7uUwqJyho/txMiBgXg0taPAxZIaQ0MMb6bSwtqcykoF4TdTOL/zEqGnIzj16wUU1bWX6wN1mejTpw8XLtwZtqVr164EBwcDMGjQIE6dOvXYRWRlZdG9e3dCQ0MxNjame/fubNu2jdat//e81fXZZWL69OmMHj2aIUOG1Ol6niaP+7XR+vd/ZtmhBXR+5s7YsVVlVQgC5KQXMrvrAvqM784HW2ZSVFTE+PHj6d93GhfPpfDlF2Px9XdFq9GiUWsxtTC+e3NdXm4Zjk6WFGQUYutiw4dDlpMQkoS5jRmmFiZsjPr2vlo2friNvSsPszr4s3vGOX7SPWp+GrWG+YOXcTs0BVNLEwztzYh+zptOUjNSl52g+ayBXNRW0s/MmZirqYyf0JQjp3ZT1XIoUgGqM6rwc7Ej7WoGLXt4czU1i7YdvbiWmsV3U4fTu01TprZ4l2qlmirfJrj5OhJZXkJnDxeiLicx7fU+vDCxG0sX7CH0VhoOGdlUl1ejqFLQrJ0PX53+190ZvHZ+vp+tH+3kq9NLCOzbEoCKkkqkBlIUVUos7cyRyRv2DVh/5UGz27RwB3u+/gPfyQNIySjB1s4crVZHWVk1ns2dCC8tpotcTtSWC7R7fQi34vPwb+VGUkIeytH2KBQaCtftx9HZHlNZAH69vblYkkNPTw9uhaXj196N6Iw8rAItKVFWozBT4mVuTUJZEV3M3LkVlkOnDu5cLEuhk4M7YUU5NLGxILe6AmmmMW6WFqQUl9LWzYWQjCz6WrtxPTaLvmkK4k9EYL5oCFnllWgtJbhL5NR8eha/QW2JLtMwbmwnXn97EKtnbuTYxjP8cGsFXi3c6zGFR/eg+b3Zbi4luaVsTVhNQUYh09vNY9gbg5i5+jXycstwcrbiy0//4MK5WFxcrCgqqaLCWY6zkxXx1SUEaiUcWv0lz/UaR/qVLFr1bUlsUDzCL/3JVFXQdFYY1oGuBL/gTDsnZ66VZNLTyYug1HS6t3IktCydlikeZOaVo2inxcHMCCyT8TL04WxGMT2dvbhalEQnmR830nPw87ImubwE53gzTIxkZGsqCXB0ICYhl8DO7pyvSGVknhtx51JxfMaLxPxinFRyZFIpiphivJo7ElZWQhdjY6I2naPriA5cP3qL7y4vw7+z7519V1SBzFiGiZlxfcV1n9pqu1SVVyPohPtuBN93OpwVW0/TIjeHP85u44V2k8i9nY+JuTHGliYUpBfi296b2OBE2r7Uj9CUImbNGcqzwwJZMPRT4m/chsk9MDSUUlxeg6+nAzHJuXRo7saNuEw6BXhyMy6Tpq2dScgromkGCCoNFadDaNLak8SQZAKm9OFqdTWBXby4npRJB1snElPysWhiQY1CjUloGS1be/DF1y9yaNNZVr/xI1O/mszY2c9SnFOKhZMVM6dtpqSwAs3lcPw6+LDizFKObz7Dymnr6TayI8GHQvj8xEe0H/C/24oP6oEaxP85Icff/ftx/Pzzz1y4cOHubG/Lli1Do9Hw8ccf3/M6pVKJUvnvTwPl5eV4eHjUeYO4qKiIoUOHcuPGjTpbx9OgtvOrKKnEwubfY9dObvo2CqWGCr8mtPB3ZuXaKUgkEoYPH87MmTMZOnQoi0d9ScjRWzh62lNZWoWiUoF/F19WXviE82djWL50P52cTQnZe4U2vQOIvBzLhohvcPZ2/J8jC/x3LU+i2srvk3HfcHnfNTZErsTWxZpxztPw7tyUNeeXcfxcJJ9sPkWHVp7cjEqnc7GCn49/Q8/5SylQCdhFKnDwtiHSuIouto7cPptK245NCAtJpVl/HyJu57BrxRScbczp8+EPOFlaUJBcQhsPJxKD0mjnak7Y7su0G9CasHNReAV6kRmdiZGpHAtnG3IdHejRszlLl4+5W+9/ZzvO5XXkxnIKs4rp9Exblh38sHZ2cB16nOyWT1jJxb3BrA1dgVdzNwQEMnJKeGXxr7SxMOP2+jO0ndSXW6mleD/blJDsHNz97UiqKMFUAfHrv8Z15hy6yRzJPJSGz/hmnNZk0cvcnZCULJp52ZNQUoyNRo6JsSFZ6ipaujlyqyiXPhYehIZl0i7QjeupmTRt4Uh8UQF2mCKVSsg3LyPA1oGo8mwGJViStOwCbd4cSkR0Dh5d3YnPLcFdYYBOq2PH7llEX4plXv+PCRzYhvDTEbyzbhojpt8/8khD86j5qZRqJri9gbWDJZtjV93dzj8O3OT7lcdZsPg5+g9qyXsDPyHuSjySV3piIDOgqLQKn5ZOHNyxgg7eHTEs8SagZxNupOXg1s2B6KoSHKWmUKOlslCBv70deWcy8J7QhFMmmXR38uJKXhod3R2IKc3BvtSRKp2acqMqWsrtyTtShH9vDy7UZNHNx42g7Aw6+9gTUZaJW5ELxZU11Mg0+FjYkJZUTHtvN4Lzs+jh7s71+Ex6yJ2JTszBzcmagqIKTqx/m/S4LN5uP4+A0V2Jyqth8qRuvPJG//uO3xEWk3Bt5sz6W1/XaWb/qa7aLi83m4myWslv2Rvue+7HBdtZ8MVshgWOpiC6jI1RK7F0tmbgRxtoJpFRuPo8rUd1JaJQQbv2XtwKTSPQWEfk2UhcfZwozCtFp9Xh0qYJ6XIzWrpbEr3jAoGvDSYsPp9WgR5ER2Zi28mF3LwyzLIVONmbk5lbRoCfM9FxuXhPbMopSSZ9tJ5ExGXR4VoeWVEZyExkOLjY8lPY14TdSmP+eztoF+jOrfBM2lsbEnbkJlbfDCehqASfWxrMTeRs3TmD6CvxvN93Cc/PHsZbK6fU+vvuAzWIXVxceOaZZxAEgePHj9/994kTJ8jOfvw+HJ9//jlVVVUsX74cgI0bN3L9+nV++umne163dOnS+xrJAEOHDkUmk7F9+3YWLlxIWloarVq1YubMmUyfPh2AadOmoVar2bp1KwCbN2/myy+/JD4+nmbNmrF48eK74ypPnjwZU1NT1q9fj1KpZPLkyXzyySf4+/vj4eHBihUrmDBhAgDjxo3DxcWFVatWAfDtt9+yc+dOrl+/jqOjI+vXr2f06NEAPPfcczRv3pyvvvoKgC+++IIjR45w6dIlLC0t2bZtG88//zwajYYhQ4bQuXNnli1bdnfbg4KCOHXqFEZGRuzZs4eJEydSWVlJnz59GDx4MIsWLQLgww8/JCoqisOHDwPwxx9/8Oqrr1JUVES3bt0YM2YM77//PgDvvfce6enp7N27F7jTV3rOnDlkZWXRrl07XnvtNWbOnAnAjBkzGDp06CPn/KD5XT55FU0B7DzzCwuXfPhA+VlkO7Ll8E80a+/NsyPG4ORky5692ykqKkIqldK3b1/Cw8Oxt7Fn2oS3+Oirhei0Onp37sflHTfINkymvLCC11+ZjdKigD8OnkShNGTNlyv4YftXYn4Pkd9/Hn+TX3yZ+Bu3ef3112nTN4CtW7eiqlERIO3I6dDjVJprsTSxw6/ai0jDmxTKoUuHAZTdriA38ypFebmYuzpi6uxNVXE25ua2tPUZxvXwbVRpNAT26EuOsZSqq5epqlYxfOAUQq+fw8Vdh8zUAjODHlwN3UKZkYY27XpQgRll545RXVhB/4lvciv+JmVFKUhlRvRTdCSo7Aw1FdV07dGXco0dVapb2NmZ35OfulrDum9/YMbst9BJdQweOqhe8vurqePrKrvuXYfx5pvTsTGVYVvuiE07H27FXMbAQk7/59/CRhnLpSshlBZIGPD8BH47tA5LAzkWfu1xc7Aj9NxRZIIEt6ZtiQo5hp2bO4LMgmcnvsHeDV9gbizH1K81LWWuBJ89iM5Rjt+YCaTHXkKVmIlOMGPoO+9y4MtlWJjKkHdpRlO3llz/bS/GhoZ4jB1F0a0Iqm8no0XGsA/mcXDZciwkUiQBfrT1bcvFfTuxylfQyrQ9id2l1MTHI+RrebH/NHbu/gYjCxlmSmsGTprAzv0bMZcI+Bm14pm5fTh27FitZqePY+/P975n+g5HIoWjZ+/8/7h582Y+/fQzzpy+ikphzAtvzmDz2uWYlNfgZe6HgWBIsiqBSq0aQzlYm/uglBahNDPnmZfe5OCPX2IiNcDJpz0SLxuygs6i0GkYNOZ1Qo+dRKXLRmFrwOB33+PQos8wMTLAomNLHNVeJAcdo9pITZ+XXiXi1GU0eelUS6WMWDSfg8uWYSI1wNTfH1dnXxJP/IGuSkuPURMJvx2KKjOJKp2OF19fxP5VX2Ck1WBt2QQfn0BuhR5AVVhG907PEGFchiItnupqFfMnfMCaNcswspZhqrRk04kfmPPOHAxkBiz618IGnd/UKa8RHRTH4B7PkHDrNlqPKuzd7e5573OwdmT6qzNY9u1SAAb0GMS+b46idCtDLVEi0Urp1bsXh4+cRya3pK9NF47fPoBPex/auLbh1uFoskwzqFRpGdXpeS5fOE6VvAypyoB+Ps9wOu0Pmrb1JsC3A9EHYkg1TaFMJmHwmKnc+L/27js6iupt4Ph3d9N77z0BUknoJfTeVTqKCgqIDctPsYsNCwqooCAoRRSVohTpvXdCAgQS0nuvm2y23vePvEYiLQEEIvM5xyOZnTvt2Zl5duaW4wfQn49HVOnp9tzb7Fn7FTXoCApvg87OjaK9m6jJKaXrlMnEJJynJjsJjV6w7ttlvP/x65xJyiQsOAqTuBoS9WepLquiT8ehxJ+Jo9Smghqdnv6T32DPmm+p1lURGtmKb9+bzpSJT2Flb8n016bf9vg1KCH+q7rE1XTv3v2GK7mRmTNnUlNTU5c8LF68mFOnTrFw4cJ6892pJ8QbF2wjOTadZ7+eyMOPPMzbb7/NmjVr6rYPaivq7/nlIJNnPYqju/1tW/d/WUPid2zTKdZ+uYmASF8mfDD2uqODHY1LY8vBeJ4d25X4HXEUZBQx+tUH6j4vKChg3LhxtGvXjk8//RSA7cv3cnZ/PI5eDtQoa6iuUJGXVohWo8XexY6H3xxOUCt/zp/N4vixJEaN6YiVtRkGg2D+l9vw8nZg+Kj2/9IRurc19vwrLShn/nPfIwQ8/NZwKvUytmw6wxOTe3DpSAKbvtuB1smOnqM7kXcikbPZpajbumAo0WCjkhN/8QjK1PN4RQTiEtYPtUYHAiyqDeirdeBsRpVcj8zJCFGoxVIoEAU1mBgp8PKyR5GRR1qVjgo/K/SeJijSa3DytKe/nyd7fzxAkZMNCrkBtaoambEcy/MF6LU6jC3NUTvZY2priSKviCnvjcIvzPtOHurbrrGxy8kuZfmCnRQci8e/uQfhwzqwY+1xlAY1clcH2jbzIjOjCLVaR49hLdmemkJBehnGlibU6HRYamQcXL0En4CWZJcmETJ2LBY6BdVyHRYyI/SlWuT2xlCswbpSRomXHJNqA8JBh1quwFRjQpVCh6m5EdVmamxlAhIMaLzN0Gv06J0EMkCvkmGmM0aj0WFqbEQ1WizkxuhLdRiZGWF2Ig+zMi1lPTwwLtZil2NAk5KNo7kJvR7uwt71Jyk0NqNZmBfNncwpzipmxEuD+faFpfiF+/DAszefBN1Ot3rvU6k0fPPVdlpG+dBvQEsAkpPy2bs7Hv9wd7ZsPIH+aBJyJ1vkMhnd+ofxv1mvYeLgTr/ej1FmLTA1MUYUqZHbGiMr0aJ1NkFtB3phQGGuQKPWY1+uoNJWg0mAoLxChoXGGINah8xDjyJXjlatAGeQFQhMFUZU2qgx15tSLdNhVWOEoVyHzkmBrFpgUiowr5GjMtajsDaCvBoMPuboS3XYVsmgsAZjpQqdSo3MYEBRoaQixAUCHdGotVjXKCCpGPPcYjRVNUR0DeHRGaPuShWnm4mfqqqG1Z9vwC/cmxNbz1CUU8KTMx8mqJX/VecvzS/j6+e+R1WhorCqgEyTS7Tv3Jbuzfuy/c8YsDBDpOeCTEaXhzrQY2xnlr6/mvQqPaYO1tQUlmNWpUJjpMBIr8cAyBQKovpHkrY9hspKFdVeTohmDuhL1cicTZElFGGiqkEZ7IZZSgkGZ3PUNhYYmytQl1bhFF9Elas9MicT9BnFyF2sCSrTYegYREZOCQM7tCBv93kCW/lxaHssGXod1qU1qJ0cCIvywaW5Myq1lmojA15OtkR5uPH9d7vpPyiSNm3/Pg5/freDS6eSeW7+k7cUX6OGzHQ7kt7r8fLy4sCBA3V/5+Xl4eHhccV8pqammJr++8OnlhVUUJRdTPyFC3Tv3p2Q4FCCrEJIPZuOf4QvAFVlVRRmFaNVSw2pGqoh8assrUKv0/PQ84PqJcMajY4d284SEupBQKArBZlF7PtxH+nlSnbU6Bj17MC66gw6nZ7tW+M4dnQrQ9oPp/uwaABSsorYfyGT4rMZ5KQWoLW2QGj0hIR5kxSTgomZCZa2FgCERXgRFlG/PmFBfjn5cSmE+TnQol3Q7Tw0TUJjzz+dRkdpfhmufi5oajTs+fUIacUqtm+NI9zXHhDolNU4OltT3qM5yjWncT6RT66lETIXB1LzEnnh2dfwt5Wzc9clSnxr+wk2uJmgTVPhXGVMsa4GZ2Nr8pUabC0tKNfrkNkac+nYJWRZ+ZgEeqDS6nAoMkWZX4OVnzGOPs6YNPdGezYDi4oKdKVKjIwVeIR6k5deSLOOzTlfUE1YoBMxsZdQ38ZGG3dLQ2OnUqrYuWI/rfq0pFfnQH7dH8fQqf3wj/AleVcscQdS0JdriUsppFJrwNvRkiA7KyxCQ1hw7iBWxlCEFplGjUJmgpdnGDbNfKg2aLDRmVNdXYV7ag051iZY+NpTItciAk3QpOkwLzGgNxGoFDpMbE1RVmuwMhhTVaHFrcQKZWYZ5mYWKJUajAWoHWSoNHpMLYyoUmpwrDam0EiNnZkpJUotjgojKjzsEK5GKAoNiHwNsrxytHmldHphMM3bBnJ0TzyGaj0PDG9LSWoecuDgH8e5FJN6RU8zd9Ot3vv0egP5uWUcSc7Gw8GC5KwymrVww8nZGidLc8xK1JRUqzGRK8DdkeSKLIZ0HUKy3BKZmQKlqho7vTHlxTXYKIwpN2jR2hiDWqDW67AyN6LKoMHGxpLqYi2OpebkuqqwkZtSXqHDzE1gsBRUawXWSlOUpSqs7U2okldj4yBHla7BrcCYkmotCksFKpkeodMjczalvFKLV7kxlTlqbMzNyauuwUVlQkVeGcalZWhNTGp3MsSFCkdT3M/nU2ECrhWCkgotVsYKVHIFdi0DkCsUtykijXMz8TO3NOOx90YD4OjhwPev/4RKee0hi7UaHZXFSlp2C+Wb77fz4MDhPDhmIOd3XEJeXombnRlKVzs01WrKCsoxMTPB2tGG6uJCLORyVAoFFv7uKLR6PH0cSb6Uh0EviN16hoqMfCz6hlGdXYVrbB5FQoGdzIEivQGFpQnVRgZkyipMLBRojI1pHeTJYV0m6k7+1JwvwTW7hsLMcqyNTEg6k4FercbSyAivruEM+LD2bbtZc3e+//0IwV08OL41Djt1DUO6haEw+jtmWZklFBVWUqWsIS+tgGObTtNnfFfKCysozC5BGG6tF+F7oh/i7OxsoqOjOX78OJaWlnTq1Invv/+e9u2v/ySuMRXTK0uVKEurcA9o2HC9CQkJVFZW0rZtW3JT8mnVoi3vTfuIqbMfb/B+Sa6vMfGLOZLIjPfX07Frc16ZPpifP1jN6i82Et4lmIQTSbz3+3R8Qz0xGAR5OaVMemom3Vq34fzPhygITObI6UN8tmALO8+kEp5cRH5iDibdWuHgYMnXCyawa+UB5k5eyOMfjL1mP8cVxZVMaDGN8Ohg3vr1xStGubvfNCR+5w8nML1PbZ3NMzviaPlYL86nlRIiU5N6Jg2LXq2RmxlxyVlHmMGY7G/3Ej66K1uO76X/4AeIP5NJfMwPtCKa6te6IJdBiaqGYEt7CrZkEtzJlxPpubQK8yY2PotgP1dScorxPJKMzCAo9/fBrZkj58vKiAj14GhODj0sXYk/mkawsZ6Uo4k4eTuirdGyLOFrMjOKcXKyxtzC5A4fzTvrWrG7dDqFl7u9y7g3h/Pwm8MBqKlWk59eiG+IF1Nbv4qqSo2yREmzNv6cP5hAn8e68fTciXz7ywH+2HcW51aWxJ4/Qxv7tqjVWvbumMeAZ98g70QhrS3MufDDHrxn92evSQnt/Jy4UJWLf4I7FVU1lLSswtvUnksZFbSzcycmIYeO3l6cjs+iZaAHF9LzCTCxobhUSVGkDDdrK1JUJbRXuJG0J4fQbj4cKMqig7MHZ7PyCdPakl9YicZRgYtBRvW3B2g3pC3TlzzD7p/28/Wz3/P03An0eawH7zzwKRnnMjC1MMXB3Z6vDn50t8JzQ9c79zQ1GnKS8694q5GVmMMzbV+j7dhuHEkppmu3YA7uT6Bt+wDOxKTjUVxEaW4p+b4meLu4Ub7rIgl2sQQ0e5iw7oEcT86mZbAnF5LycIxyILdSibESnBwsuUQ5kfauxJ/LoYuDGxf2pOD2RCAny/MI8LIlq6oc20wzbCzMyNBU0NLamfgLefTo78sOTRzRGS04fzoPt3bOZJSUY2JnjJmJEbk6JRGmTmTszyMqzIuYs5l08HTjwskMAg0qsmLTsezTFkzkpLkLWllYkzJrK6F9wjm7LY6W/Vpybtc5mo3vTVpmCV/MGk3gPfC259/sEOCXX35BkyDnt8/Wo4zMZv/Rq7/h3/3LQeZMWkDEyGjOZFXQuUtzjh1NYv7Cifj5O/PEo98hNFrKdp7Eq1cIx3ws6KI24uIPB4ieNpTDMZk0Q03G6RT0z3VFAGWqGiK83Zg9aSiHEtJ5fflmutm6cfFgKr6dfbmUU4TvsRR0ah1VZVUERPrx3h+vYm5phsFgICUuHZ9gTz599GtO7zzLp9vextHPFSMjOVVFFTh62JObUoBXC3eWz1jFuq8388nWt2nZLfS2HLt7IiGG2m7XZs6ciVarZdKkSUyfPv2GZRrzpXq+05sknkxmY+UKTMyuf7PTaDQMHTqUH374AS+v2qeEQ4YMqasTKLk9Ghq/iuJKRro8iX+/lsQ6mNPRxITzy/fTskcocXvjCWrtT3JMGvYe9uhrtOSV5JJjl0xS4UU0ai2jx4xiSPCo2tbyE/pyKamA5SufxsXVDrlCTkpcGk+3nk6/x3vw6tK/h4Q0GAxXjHZ0KSaVZ9pMp8+j3Xht+fP/2rFpCq4WP4PBAIBcLq87flv/jOGLWZuJtDUibtNJPtj4Op0Gt0Fdo0UAQwfMwsvRgqw/DhP6YAe2nz2IbaAlquZtiRb2rF75GQP7PEteQQUcisW5oz8x3VxpZ+dGQkwOEVHeHM/MpqvcgYunszB7wJOM0jJOvvssWcUVDPnmR9qbO5O1LZ3I1j7Ens4gqLs/Z1JyCZVbkplZzPpN/6OysoZHRn9Dp+hmfPjJaAwGgVwuu+r3oKm73rlXo1JjdtmPvZkPz2Xvr4f5IX4uHkHuDO77GZ4e9iz5+Wn2rTnCR6PnEPVYb2JSivEe6M8fS2YROvohSqxscE2QkbhvKWHeo2kR6k5CfC4BD/ixR1NAB2cPTqfl4O/nSFJhMWYucizkJlSfURHq7Up8RgEtW3tyqDyL9u6eHM/NJtzSmYTCIhwNpsgVcvJFNeH2zqTEF9Eq2JPjOdl0cPPkdGI2kTZOJKYW4GBnjq5aCysO4d85mGRMGTKsFS++MogFLy3l9682E/xEfy4m5rHi12dxcrZGrpDf0zG/XvxmT1rA1iW7+ebkpzRvHVjvM41ai4lpbRUTExMjamq0mJkZ88Izy4k/n4WxsZaTsSsJaz6Otm2DSEhdzaNPz2D2T3sJifbmWE42bWSOXEotxMLXAp1BUFGlppmHE4k5RbQO8OB0Sg7hndw5VJFFSxs3zhXkE5JiSY1KQ7G5nha2dhTvzaHf4AimvzaMOev3sXzXaZq3dOVMYT6+CmuKVVWowirxVziRfkpFK393jhfk0sPEg7OnswgPdONcch7NdKYUFFWi9DDF38uRnz55nG0nLjJ9xRY6ubpzJj6bdvYOXIjLwbu0mJyEHDarfkZxl54U/6WxCfG1rkH/nL59+3bmffYtNXuMeeqLx/hx7+K6tgeXz5scm8bUVq8SFt2C84cSeGLmw4x746G67wWARqNlqNWjeAW6kpGQQ2i3YOL3X+Sx90bz6LujeGvuBnadvETImUyyL+Uik4FbmA/ZVnZ07xnCO+8PZ+Guo3y95wg9qx1JOJmJT2UpWWcz2FDxI0VZxUwMfpHOD7bj/d+ns3bunyz833Iie4QRu/c8fhE+ZMRnYdqnLRZGcko2H6NF+yAuHrtEq14RxOw+ywsLn2LIlD63IyTAPZQQ34zGjnRWUaxsUF3A3bt3U1JSwsiRf7c037RpE4MHD77lbZb8rTHxS45Nw8nTgb07Ylny9GKGPt2fcW88xPznf+D45tM4PNGD/CoV377wAJ99+ilJf+TRpV8HTmyLxaGNBcUnqwjt1JyEE8n4hHlTmF6ImaUpDm52fL7rPRJOJOET4om5lTkAP324ht+/2sSXBz+qGy/+LwknkvBq4YGljcW/dmyagn/G7/2RX3DpVAom5ibYOlmTfSmX7qM68/SXE/npiw2s/Xgtj703muEvDObdbzYRl5jDqi+e4FJCLu+9vQaPMCcOOJTBL+vwuuROWJ8IEvZeQHhrMS2yxMrBCkwVpPYLxM/NgYTsQtrrTUlevI/QPhFc3HOe55c8R0SvECpr1HhYWzNx/EKcAxy4UFZGK3cXLh5LJ8LJjPObT9KsU3PST6fi6OWITq2luqIanwhf3vrlRRIu5PLBjN/p2syRQyv28snWtwnp0OxuH/Lb5lrnXnJsGq/0fI+H3xzOqFeG8dus9fz66R8ERPqSfSkXJ08HVNUalCVKWrQL5NyBCwQ91J6TQkd7NzdOHDiJyrEA09COeGKBWqPjXEUMbVtFk3e6iOBQd45W5NLa35OYtBxCAty4WFSIn9aaSlUNFfZ6vMxtyLlUSqi3C3HpebT1cefcmWxCWriRlFyAY5A9xRVVmJYZcHS2Irm6glAvV+LSc4kK9OBMUg4t/d2Jz8jH19mO/HIlzhfVWDubczpIQxulGblz9hL9UHsO/n6MCR+No/djPSjILyeomRuTwl/C3s2Oz3fOuIsRur7rXTtVShVfPPEtZw9cYFHcbOycr1714/edsXz72wG6+XgSezwNt+IiTl46jF1EW4JCw0nSKnGqSEe5JY8hrz3IuJcGM+vHXRw7lUqQwpKSwkoK2pvjZGFBXlY5EQozMr/ZS7PJPTldUklgZw/O5Rbg72JPbkkl9nkyrCxrE1j/Zk4co5AH1XacmrUN79cGEFteRlRcMQVns1BN6Yi5sQL90WK8Qlw5V1RClKUpiUv3E/Rcf84WlBGWXU5ubDof7ZyBkY0ZVhamONtbIYTg50XbWDX9Z8J6R3Bh73kCx/YkI7sM0dYOYzNjfnn1kTsRpmtqzL3v7IELvD30E5764nEGTepdN/294bNIikljedK8ugT/lVdeYcaMGaTHZRPcPoht27cxePBgPho7hwtHL7E04eu6qoXbf9zLN9OWEBYdzLmDF3nrlxfpMKh1vXWnX8jCys6SM3vOMe/Z7wkd2p6zBdV08bfn6KqDPP/zNFq29OOtQR8jhODjzW9SotTg4WnP2T3n+GT814RO7MWFhCKCmruRnlqIY1YOBr2eqtJqXP2cSTufyYgXhzD29QdJOJGMX7g3sz/ZSOy5bJydrKiq1qA9fBafYE/e/u0lshJz8QvzIiUug5COzW7rj5v7JiFuqKNHj+Ls7ExgYP1f1qtXr2bUqFG3ZR2SWjcTv/KiCrYu2UOX4e1JPZvB6tkbMLcyp9zSBFmUD+VZZwkstAEDDH2mP6mx6aQpk/C1DCKolR9/LtxBUU4JZhYm6LR6hBB0GNSGcW88RNKZVH5442eieoZzZMMJTPxd0Id6EFRcjbJEyatLnr3mdm35YRdHNp7klSXPYONgfbsO0T3tr/i93PcdIqPDOb0zjh5jOmNuZY6ZpRmleWU0axOAqbkJ3760FDMLU7QaHS8smEKOWsNvC7bhWqVB7etBRVUNedYa8vLO4kcLnMtUlOkFtgoZZ9LOEG7tj1alQd3cFRHujtKgwdrSjLKyCjxTlBSaK3Ar1iJ3dsSzrRcFZUp0VnKUaeWUm+iwNjenXKnCTWNCaXEZTlklKPPLMG3hQZVCjllyPsZGRmg1Wno/3JUuY6LZvvUszb3tuXQ0gQFP9qqL65dTF2FmYcLUORPubgBuwbXOvaryKn76aC0XjybSYWhtP5/m7o5UCBmWFZWoq9QMe6Y/1eXVOPs4cfbIJU5nlmAa7kpyYQrWFVoI9cOhyghDkRqZQkZi3kl8fVpjbWFOnqUGOysLlHlV2FqbocqvQhZohbpMi8FZIDORYbYtCzNvZ0qsjbCxN6fARIl7lSWF1OBcbUoFWsz1ChRqgVarx8TBlHKtGl9Xex5sH8bGI+fJyCjBtkaDOqMYIzsLTIt1VAQ7YBpkzcORYZz58Sjn04oJ9rLj0dcfxP6y+sJbl+5m65LdddeFe9GNrp3nDl3kt8/WYWJmzJsrX6xXD/MvGbklLP3tMPkXC7G2MOXCyX08MmIg1oG+2NhZcD6viJW/raSjUQCanDLaP9qLQ+cyMbEzQ5tfhczWlGpXObK4bMyd7VDH5+BokFOq0mDeM4QimR4boUWdXISuqw+ycgOm2RpMbc0oMtNh7m5GVWoZntnV5HpbY+1ojToxH4c8DZVBjqjsjbCvhHK0eLo58FDr5uz67SDnrGQ4WFqjSs7HUmPApo0fZkeTcfRwICsxhxbtgojZFYeJvTXlCmNaBjkTNbgt65fupURfhUwup727CxM/GncnQnVVjbn3XX7P8wxyr5t+fEsMRVnFDJrcByEEv/76K+PG1d+nv/KWUztiyUstYNDkPnUDiynLqljxwWrOHrxA89YBPDpjNEc3nuT4lhheXfpsXTeEZYXlfPb4fCxtLBj03ED27DxPyo4YIqNb8MjbI7C0sWDfqsMYDIKeY2vb62SkF/HVpxvwUEDPR7qyeds58nJL6dy1BR4mMnRqHeVF5exddRgrOytUldWMe3MEHQa1ZvfKA2z46SCmLnbUlCoZMLkPhpwiXHyd8Wrmzrznf2DIU33pPKzd7QwJ0MCR6pqaouxifv5oLXlpBY0qV1ZWxrRnXuTgLyfQ/KOx3IoVKxq0jBEjRpCWltao9V6t/JEjR+jQoQM9e/YkNTX1qvNWVFTg6elJjx496N27d4PKNAXnDl1k9Rcb0NRogNpeP/atOcpvvxwhO6+CMdMfwDPIHblCXvvrUAhMtFB4MZ7E86cwtzAjsnsYzVsHoNXo2LhrE1XNXFE7W6MOccfJz4nAVv6ow93QtHBGYVR7GsjlcnQ6PbF7z2FmZUbYkDaYWZqhMDaqG6jhWhRGCoxMjO7LUQyFECiMFMjkMnRaPXqdAb1Oh1ajo6Kimi07zqOR1R7jmqoatizeibqwCuNqA3nVOgqLKtFaKYjdvx6ZSqAwN0Ud6oHM0Rbh6UxC7hnwqL0YurcJxDw2B7PMCszztJgII4SnC7jYoAl2pxw9CekFFKWXUZJUgiy7EHm5Ep2ZDJleBxl5yI0V4OWK3NgYmUGgcLJD5uYIHk4YbK04tTOOgpR8xjzciVZdgxn96gP1fuQYGStQGDeoPXKTY2lryYAnemHlYIWRkQK5Qo4wGHDwcyW4fTMUCjk6bW1sjYyNGPhUX6xc7NDptJxevwJLK1tMlQYsirUY1xgwUelJPbgf81KBAIy0YJmhwTy3Gqsj2ZiWGTArFlikFGMRX4BRlUBhUKC1loGpHL2JABMZalsZckXtueVuZ0NzZ0fUDjK0tnK0ViCTg3muFqp0yKsNyGUg5DJMqwRmpQaQyVAgp6OHN+FeHujdnJCbm6LXGRD/X83n2KZTrJu/Bb1Wj0wuw8j47r5WvxXh0cFEdA2hNL+crUt3s27+Fo5tPl1vHh93B3pGBWFWXUPJxfMk5J1i2JSBqMxkCAsjhvWKIOPicYSnCworCxRGcoz1YFypw9nFFr8QN0wqZRhr5ejNQGZlivByRW5lhgDMqg1YZuuwKBNY5ugxTSnEJLME44s5GBdVYaQEuZkZ6hbuGFdoMD+VhV0Ld5p1D8cisQizlGJ0VkZgosC4TIsQcgwezhgrtZicz0Y426Br5kLxyRTyMwpJikmlrLCccwcvoJKBOtQDvbsdqhbuhEb5EObvSADGBGKM0VXO39RzGaz8+HcqS5V3KEoNY+tkU3fPu1z7ga0YNLm2usDSpUtJT0+v+ywnOY+fP1rLD4trx3Zo0zeSwVP61rs/WdnVnuuO7vb0fqQbju72KIwUVJQo+fO7Hej1egBkMhnGxgqEENjZmNOrW3NM7SzRuzsh/r8Khs7OBr1DbWJfVVHN5gXbEGotPSb0xM7FtrasnSWDhkTRY3Rnej3SBZ1Oj5GJERHdQrBxskGukJN2PpMjG09hKReE+jrgbmtK3/4R6LR6VJUqNi3aQU1VDQrFtVPXjQu2cXjDzY0Z8Z+8qqeezWDtl3/iE+KJWyPGoF+9+ndauUWzZeFuhk7qj4Nbw7tT02g0jBkz5qYH7/hn+ZkzZ7Jlyxby8vL49NNP+e67764oEx8fz9SpU3nnnXeA2nrONyrTFBzdeJKdP+2n58NdcHS3Z+2Xm7DxcCDXxpbU+Cy8PAaQGpdB+4GtSDiexObvd+E+pCMXDu3k4JHV2NnVPu2JP5rI719uIteqmF+2nCTcz42LReW88UQvuvWNYtSnK3CwtmD0sw9w7tBF/MO9GTKlH189vYixH45j5Kho5PKGJbj9Hu9Bv8d7/ItH5d41/IXBdBnSkbRzGWz8dttlVSbyiBjUhpM5SqLbBnFo6S46DW3L9h/3kSkzJr1Cg7mZKTYKOYm5ech1emrcmmHibk3c+Rza2NhxKTEfYxMFmZVanvhwBDqDnpmfrsezfXNSk6tpFulJYkwm7QeFsi8uhSBXe9JySnCvrk3cyg/E49Lal1iZoHWVjKSNpwkd3pkLueUE2luRsfs8Dn3bUqYwoaZah6ujLee2nuTPZXvxCfPG3NyEqvIqUs9m4BbsRWlJFc/Ne/JuH/J/lW+IV91gI0LAqlnrmPXeKCxtzDmx+TQbv91GWUEFrXpH8NKiqUR0C+Kr736mbXhvShLURLWx4FxKJnaOVmhUWlQ1WuQmCi5mFBLu7kTSyUxCHcy4uOMMXs/1JTmtCN/YXGrKlOSPsMSlnQcJulIiTeyJS8+ntZsbMfn5tFU4kZJeyIxx3XBu4c6fa/7A1dacRGURrXXWnNp3iUqVmpjsfJqHuHFOVUm4hzN5RUpkcjnOeWoGBftzYHMMe3adJ9TFggMr99PzoXbIFXL+mLeZzAs5mJgb4+rnwqhXht3lSNya0a8+wKmdcayfv5WaKjWeQW60aBeInbMtWo2W+MOJRHcJ5sKfx5m/bjuzv/0QU1MTftl8imY+znzcaihOpuYkXihgaN8oBgxrTUpyAafOZvPijGH4Wpkz7VQSdkEuJJkqiWzrx8W9WXTsH8Wus8lEWdmSnFeJh58rFRdKMT2XipWLDVmJubj3CeOkIp++wQEcOJxEpFZP2pY4vnxjFHKFnOPf/ImxgyXpKAi0sSLubDKakhounM8m2t+Oo1tjcZ3ci8z8cjwv5FKt0ZKXWohXc3cST6XQfmpfduWX0K1NEFuOXKB5gAuPvDUCgNSUAiwsTIk/mohHoCvp8Vm0aBdE7N7z/P7ln7Qf2KpJDbJkMBg4c+YMc+f+PYpqwolk1n75J6rmquuW9Q/3qTew0IAnehF3IJ5N323nwecHkp6UhZW9JRM+HMdrfT8grFMLhr84mKSCKn5beYQBD7TBysqMDetPIQyCPv0iKMoqZtfS3Yx6ZRht2vrzx5oTxJxKY+qEaOz+/6mzVq3lz4U7iOgawsNvDK9b/8YF2zi9M46Pt7xFi7a1b+mT49JZ9fl6/Fv6kngiiaFT+9O6b0ti954nvEtwvbcfQgjWzP2TgJa+N/UE+T9bZUII0aindTt27ODIvgL270/jm0UTaRFc2+2bRq3FyFhBXk4+Hl7u1yxfWVlJTEwMS5Ys4b333sPPz69R+/LP8k8++SS7du0Crj0i4A8//MCSJUtQKBRMnDiRn3766YZl7iUNjV95UQWj3SYR2CaASyeSieweRuy+87y4cAqDp/Tl9QEfsnn7JvbmbsHKzgqFkbyuXtG2ZXv4ZOKXtJkwkNjkYkJkGhIOXmBl5kJsHK1QGClIO5fJ1Fav0veJXkz//mmWLN7DyhWH+fjzsbTvUHtS6rQ6ZDIZOq3uvu9d4i9Xi99flxOZTFYXQ1VVDeaWZlf8PTHkBYpzS1mW/BW/rdvIwthiWnm7EpOZT69yOLfiMO+sfYWwbkG8+sIqsvPKUbqb4OliS0ZeGcFR7hwvz+fp7u15vm80H47/kv0rD+E9rhfZOaUUdrPCTWeE9tO9hHRsxoUjiUT2iiB291nCnujHucQCfJVlZJ3LxNTSBBsHawrSCgnu34p4pYFx4zvz5JSefDl1EZsW7SDw8b6kpBWxftMrKORgZGJ0Tze8up6GvLKtqVZjZmFaF6+prV+t/dGj/Akj49q3IXsOX+D5DxYS1dyfpGI9Hz0/mD4dgxn32U8k5hbjcUGDqbmeshI9zVq6E1NaSkdPN84dTcN3cAAns/MIs3QgPbMIubsppqZGpDpV0aHUlMIP9xHxSHfi0suJaO3N2dOZBNZUkno6Bd0rPZHrBbLjpQR4WJPy2wFajulGbHYFLaO8iTuTSVCwGymJ+WgHOKGrUmP+yQECovxJPp1C+IjOnM2v5rkX+vHgiHa888CnHP3zNM7ejqgqqvmjZPkdjkjjNPSVu16vR6/TU15YwSO+z9BhSGs+XP866+Zv4ZtpS5i+7DkqLIvYO+ckCceSWZW7GFsnG2QyGaeOJvHyS0tp7+vCmVUH6xo7Ob03hIvFpTRfG49OQEWQL0HN3biUkMegYa14+ZVBfPPSUtZ9tRnn2UOJKy5m67THcbGzQaGQk3wpj2emLKXfoJZMf30oc7cfZPH+E3wzbig9QgN5IvgFivPLKR4Wha+ZKYXf7yF0eCfOF6iY8GQ3xj/elQ/eXcuBvRdYtHQynl72DLN+FI8gNzIv5tDr4S688dMLddebt+euZ+fJJH79fAJ+Ho4M6vMZTpbG5G48QniX2vqzj747isfeG93ovOHfjt+NaLVa1q9fX6+9019UVTWo1NU4ODg0eHl/nfNCCAwGA0MsHsG/pS/fnvis7thcPs+1jtU/P/tswjx2/rifxWfn1LXjulb5v6arVWpMzU0Z7jgBCxsLVqR8A9TeW9Z/s5X5z//ACwsmM+SpfqjVWoyNFei0eoxv4U3tfzYhboz09HQe6P8QrqnNeG3lS/Qa0REAvU7PGI/JeLXwYMPRtbzy7HSe+XLidZc1YcKEm0qI/1l+4sSJ7NmzB4CePXvW/ftyu3btwszMjHbt2tGvX7/aG9QNytxLGhO/3NR87Fxs2b/6CF8+9R2jpz/AxA9r60stX7qcMzGxzP16DmO9puAe4Mrc/X8PojKw8xA4ZcWz302lz+hOlOaV4R7gynj/Z7B3tWXe0U/YuSuW93/czVOjujBuYBuyMkvw9XOqW8aEFtMwMTOmIKOIDoNb88ZPL/w7B6UJaUj8zh26yPTe7zPxo3GMemUY37/+E+vmbeGbE59i72bH008t4+TplTjn29KuY0cSDiUy+KXBTP1wLPM+38z2nfEUJG3Au6IZ8se6IJfJ0CSU4tnSlcOmpXRy8STuWAbd2wRx8HQyHz7RlzatA5n2zDLUWj1VFTX4OJiT/PtBwh/sxJlcJa2jvImJyyKgix8JWYX4aoxR12hZ8N0ElKVVOHrYk5dfiZe3Q10VgbzUAhy8HKmsqMHJyYoxHpMJah3Ap1vfvsNH/fa4Uez2/naITx+dR/SD7Tj65ym+OjQTN38XpnV+C5kMfjj/JcmxaTzd8VUSLE8TVB7J5DmTGPv8ID79aD2HD11i9sLHsDAyonuXhxjYZSypvx8kdHgnYrMqaTEwiMO5OUT6exCfmU9YkQmVZdVoD8bh1tyNjNgMmo9oz3GhJ7KjDwfysuh9TkXyzvO4P9iForIqOHoOhyhvjvVzohOOpG9Op1VrP2JOp9G8ewBx6fnMfelBfAOceOqRRdjbWpCbV0ZzF0viVx1gxGsP8eT7o+tu8m9O/5XU1CK+++5xXLycrnLU7h0NvXY+3/ENygrKWZHyLbkp+di72dUlMxkXssguy+TFCf/DLT8I9wBXygsr+DVrEbkp+UyJ/B+lvrnsO7ODH9//jVWz1tPspUHEFZQx9/Xh+NlZMzniZdwC3fh85wz2Hktk5tId9Anx5+iOCzz/TC96DWvNm4NmkhmXgXG3KJydbfh6wQR2Hojn/e+3M3V0NGMHtuGTpTvYcfAibbPLSTmVwrcnPqNSDk9/tIpAX0dOlxTSzcGd87uTefu9h+jQMYh3h88i8Wgits42yORyvj48k/KiCpy9neoajq34YDW/fPIHr6+fTrd+UQBkpuYzrc1rtGgfxAfrppObWoB3C487+uP2duUuM2bMoEWLFjz88MP1pm/5YRdfP7MYu4EKfln3c4OWteStlayd+yfzj3+Kf7gPAIVZxVhYm2FpW/tkd8+vh/jssXm88dM0uo/u3ODt1Ki1FGQU4dXs2g8VL/fFE99y4PejLL34FTqtjqdbT6dFuyA+3lw7GqgQgh/e/Jm1c/6k7bNDOXM+B9+KUgozivgte9ENln5t/8kqE5fbtGgHO1bs4+1fX8LJ0/Gq86xatYrvFi+kOKGSjv0i66YrjBRM+3YyNk7WHH1pN4Of6nunNrveL5xrnagdO3bE3NwcuVxOx44dOX78+A3LNFXu/q7kpuSzccE2WvVpybFNp0FAzO6zxCnO0SpyOO99sxmHfi1Rn81g/rQfuHQ6lUGTeiOs9EybP4luD7Zj3bwtnNx2BhdfJxw87NBp9Pzwxs+MmzGK8Xnl7FsfS7CLI23bB9Rb/9NzHkdhpKC8qBLfUK9rbKVk9md/UlWt4d33a1+DBUb68tz8SbTtVzs61oAnepFbVsPc+bv48LMxDB8TRnKMFT16dGfcm8M5sTWGIxtPssLVgfjEPEK6+hOXWkmPQW2oNBgx4YnuFCUXsmHeJrr3CKDsbDGdQnxIzS4mKsqHH07GsbMqF3VLS0wqwU7mQE1aBS1HRFOg1tOqcwAZVVW0au5MwYELTH1pMP5uTvz4wWqWvPkLLy6cAoCv399vAYyMjfBqXvvGyOL/B4uZ9u1kHD0a/uSlqYnqFc7z85+kWesA2g1ohVcLD8wsTHnmy4kY9LV1br2auxMwyA2TmDZEdoxi37I9KIWC1NRCnpjUnQAvJz4Y+TlCU4xaLidsVDQ5ZjL6PdiKwUOjcFy9n7RN52jdJwyVQYW1jwXGrm0wHE8jvFc4hWoZnUO9yI8toZeHHZUZWYQMaUe5kRxvhQFZmDelfi50tnKjLLuSll2DyCivpFWHAPIzK4gMcmfx+qN4J+bhIdOhNpYT0sKN0hMXad23JSc3nkAhBDG74nh16bNMmtqbwsLKGybDq2dv5OifJ5mx9pV7vgHtEx8/TE1V7cAyl/fBH7P7HEvfWkl1UBGLly2iILkERw97VJU1qKo1fD5nG23HdeNg0moURnKGTu2Hq48zvt2CWbloBz8/u4QP103n2a8nsvrzjayetY4R0x9Ag4FgLxesFcZs2n6ezJNJqMuq8Qn1wqBVM2FKT/Lzy1n3y3E6BXiyd10s2rPZpG8+ycSn+9Ha24WfP1jFwleW8+6aV5j6YEe2zP2TIX1DGTOqK8mRQbSM8sHE1IiJ741m6Tu/oixTMuGDcVjZWdY1BPtLn/HdyMqvYM3a05xed4KC9ELkchkt2gcx8cOxmJiZ4BvSNK/nSqUSpVJ5RTIM0GFwa56fP4n5v89p8PL6T+yJR5A7nkFuddOcvernTK16114XonqFN2pbTUyN6yXD7w2fhYObPdO+nXzV+Ye/OJioXuHYOtn8fx42BWcvBzIzipn1yUaGj2zPsGcG4BHghn/HFnTNLsXZVE51eXWjtuuf/vMJsYO7Pd4tPDExM2Hz4p1UlVejVmnwau5OjzHRLFu0HLMsO1wd3Okw6cqBQLqN7ARA1z7RdSdO0plUdv10gBEvDb5mkn2rzM3NKSoqoqCg4JpPm99991369etH3759OXXqFKampjcs05RsX76XktxSDAaBs7cjrfu0xC/Mh+AOzTh/+CKOXvaUmuYz7ak3yMwpw8LVFrd+LVEZG5NxIRvz/+9WzcfTl4yL2ZzYdobzhy6i9XYgy8YMY1MTjE1lJJdrOLzrHMoDF7FWqtmxYAueziNx9//7BtJhcJu7eCSaDndPe2pUmrq/za3MGTSpN+lphSycv5NgTxu05UoUBlj01hKqDUq6+kbTflBrPILcKM4pxdLNnvikQszszVGbynDw8kUT6Ix1fC65hxLIcDfH1NcO02NZOLg5IStR4+dsS6vWAVzIKcDdzpZiXSFyg8DIAKY25uiNFLgYKejULZj9Jy6hz63E1tGe4rQyTLNKMa1WkZeqYevS3WQl5NBuYCsiu4dddR+FEGQm5KDT6Ajr3IKCjEJ+/2oz/Sf0qBvJsqmzc7Zl8JTaBwDN2/7d405OUl5dQnzw8EGee/Np1szdQbWlFS42lri629FcpaVX33A27D1LqZstDg4eWNlZUONkinWViqozyeR72MDxdCy87NBYyDC1N8NgJMNUqUbuZIUq1AMrrRxKNdjYmqNVyPCJ9CO0dyuObTyJqKgivEc4uR425BRW4CA3xVCjw8PFlk5RARzedQFtmRYhh2KFCcYVpZiYmdC+RzBpSiXBHYM4fygBFx/H2mTf0gxvr4Zdy529HPBu7oGxyb1/+2zVK+Kq020crMjWptHRKBovfy8iO0ewdekeygsraDewNb4BzoSEeqLcdQkAJ09HBk3uw6kdscgS8xA6PWvmbEQul2NpZ4GLjxOqsiryNp0hbHxXevUOw6A34GpnQot2QTi426PT6Ihq5UtZWRXurraUl1fjbG+Ns6cDzQLdGBQdwoWjl5DJZFSVVbP1+130H9mRS2uOoy7QIlMbGDKsNefOZrLq4FF8rGqPf1jnYNpe9iALIOtSLn8u3E5wnyhUtja4W5jgbiYj62I2Mhl0HNK2SY82WlFRwcGDB5k9e/ZVP3dws2fQ5D7sSdzW4GV6Brlf0XDvny6/LtwKz2Ye2Dlf+WQ84WQy+347RGCUH1kJOei0OhRGCrqPqs3DCgsq8PFxxN7eAhdvJwZN7sOJbWfIOhBP1zeHY25pdkvbde+f0beoWZsAZDIZJXml7PxpP8hApazBwc0etUU1v61ahVOqH617Rl63j+LXXnut7t8ZF7I5tO44PcZ0/tcS4nfeeYchQ4Ygk8mu2cPFyy+/zPjx4/nggw947LHHiIyMvGGZpuT0rjiyEnNBCKzsLGnRLpDOD7bDL9IPuxBv9u77nR7jOzN8bKd65Up7RfJqr/foPqoz7Qa0Ij//IX59bz0ZF7PJvJiN7cj2FGl12JZWYWZjQWFJNTs2xpC9PZbOw9py9M9TFKQX1SXEKXHpVFeqCI8Ovua2FmQWkXo2A2cvB6rKVUR0Dfk3D8096+Hx0VdMq6lWs2PNUQ4cS0PrZ0dmTArN2gbw1dLZRPv3xc3OhYhuIRxYc4SYXWfx7dOKlJwyvFo4kpRdSoe2gynVaHAvV3J8WwxJnb1w83Uk52ASHds350h8Pv1CIniwYxijTSNJySriXHY8WqUapVqHk6sN2dmlRHUKoHt0c4y1gp9iD+Dlas+RQ5ewzcrF2FiOsqianSv2U5BRhJ2L7TUTYoPBwKF1xwmM9KPXw10pzCrh8PoThHZq/p9JiK/l+JbT6HV6Wg0L5a3pb/PrT78R2L45Ww8kMe3lwQT62ONiYURxUQVb98dTY2tBUPO+yAx6MgvLCJIbkbD3PI425lyKScV2UkeyqivwrjZGpTOgPZOJnZ8zWSo1nUN8idmTiGsrN85Vl9H1wfa0D/Lg+JId1FSrmTBjFIWlSp77eDXtAj2J3XuJJyb3pGfvUJQl1ezcehb/QBdSdZUoiiuxk0HH9gGMHN8FgCFT+gEweHLjbvA9xkTTY8yV3/Om5PSlE5h5GRG3P569vx2ibb9Idv28H41Kw5jpD/DCywPRaXXoivpRVlhe15dxwolkcpLz8Y/w4difp5DJZJhbmxPeNYTinFIOrz9Os9b+tB7Ums5dmuPjao13oBu2TtbotLW9FtjZWTLm4U68/85aOnQMosfwDgx+tBsAv366joLsEkwtzDix7QwDJvWmxaiO/PHjEVJTCgkO9SQpMY+DBxIINYf89EJGv1I7sqjBIDhxLJmAQBfyUgs4+PsxMivUpJWpGfNIZ8IDnTm87jjmVmaMeGnI3Tnwt8lLL73E44/feNTcy/OWe8nkz8ZfdXrauQwOrTtOVXk18UcSGf7i4HptdpxdbHj1jb9Hki3NL2PHj/tIjknloWmDbjkh/s/XIf5y6iK2/LCLwEg/Mi5k8XvxUipKlDzq/wxVAYX8cXAVCoPRNTsvryqvwsTchBEjR9SN+CK5PRpajyovvYCJzafRol0QF44m0nJkNCezSrGwimfj5l+pLFFiYWOORqUBIyOMjRXUVKqwsrdELpczbNgwhkc+zMqPfye8azAXjiTi4utMVXk1qkoVXu2akyIzpm//CKa/MYyywnJsnWwoV6qws7bgqahXyE8v5LecRddsUPfZ4/PY88shvIM9yEspYKPyp3/rsN0zbhS/8vJqrK3NObDmCB+Nncvz8ycx7Jn+QG1/39NfXoCDXQS6fTF4h3iSEptO2EOdiMurooOnNcd/O0B4dDA/H1pCbPJp7NzskclkpCfl8fxzK+jbP4JX3xjKql+OsmjBLt54bRA9Bkby5DsrScstxvb3UzgFupFlY0+zcDdOlpTw+OB2PP1wNz57fD67ft6Pf7gPWYk5rMz8Dm2NhkcDnqXTA+14d9X/7vThvKMacu5dngj90+svv8nBr8/QqUcHzuw+VzfS1dwpC9m6dA/+4/uQkV6EsbGCmLMrCCsOI3RQG+JKNIwa0ZYp0/rz9bPf8+fCbQRG+dW+0enbDgsjOYV/HqHDkDa8t/ZVNm2MYe7nm4lq5UtcbAbBehWXTiTx46V5WNhaYGJmjLGJcb1tKy0ox87ZhsoSJeY2Fkx4ZAG6Gg2VO04S0rEZn++aQUWFCjs7y+vu473sVuqgCiF4/vnnmTNnDotf/YkN32ylebsgks+ksabgByysawcnij+SQPfOPfnktVlM+vTqCUxVeRWj3CYT2SOUT7b8XZ/+m9mbWbchhnAjHQmHLuLk7Yi6Ss1vOYvr5tm7O56P3vuD51/qzwMPta2bPunxRRTll/PbH9PIKVYy/o3l9O8YzLvPDroiXk+3mU5OUh7ry38kK7OEJx/7jm7RQbz10ShWfb6exa/9RPiEPpxPLca/ooychBzWV/x416sU3kr8MjIy+PHHH3n77Ru3Xxg2bFiTzls0NRp0Wn3dd7K8qAIbR2vKiyqwc7ZlwUvLWDdvM/OOfULzNoE3WFoDiCasvLxcAKK8vPya8+j1elGcWyLUNRpRXlRRN33J4mVCq9GK/WuPij6ykWL9N1uvKKvT6sQQq0fEi13fFkOHDv1X9uF+1pD4GQwGMdR6vHim3WtCq9GKHT/tE71kD4nhrceLvvJR4rX+H4oBJmPE021eFUOtx4sHBn0hJo2ZJwaajhWfPva1EELUxa4op0TotDpRWlAmqpUqoSxTitTEHDGg1ydi2jPLRN/uM8Xbj38r+shGipff/ll0Hj9bJKblC1VVjRjjOUVMDHnhmtup0+lESV6pqKmuERUllbf1ON2rrhe/woJyMaDXJ+Kj9/+o/Tu7uO6z2NhYceHCBaFSaURlpUqUF1cIdY1GHN4bL/p2nym++Xq7EEKIuVO/E/2MRouuHbsLIYR4ZNQ8Mf6hL8Ug83Hi9YEfCb3eULfMeS8uFf0Uo8TUt38SXR+fKzLzSkVlqVLUVNeIU3vPi35Go8Urj8wTvbt+JN4e96XopxglpkW/JQaajROPNXtOjPaYLIQQoiS/TOi0un/piN07bnTubVu2R/SRjRTbf9x7xWerflklBpiPFq/2eV/odDpRlFNS95lerxcleaVCrdaKCcHTxBjPyWLw4MG1x1WnE38u2S36yEaKjQu3C4PBIGY8NEsMMB0rUs6mC6WyRlRXq0VpQf0YfP/dbtGn20diz85zQqPWiKdavSKG2T4qhlqPF9M6v1lv21bP3lDvuvDUxEViSL/PRGWlSlw6nSwGmI4Vr46aI/p0+0i8/dg3oo9spDi07vhtOqp3TkOunVej1+vF6tWr6/42GAyiOLdEaNQaUVZ45bL69x0gDAbDFdMvV1ZYLrQabd3fv3+1SfSRjRRbV+yrf70tr7qibGFhxRXTVCqNGO0xWTwZ9qIQQoiDhxJqr80Taq/Nu1YeqJu3prpGVJYq6/4+viNW9FOMEt+//pMQQoivn/te9FOMEse2nxFqlfqeuTbfbPwyMzPFoUOHGjx/U89bpkT+T4x0eUIIIURafKbobzxGvDHgI9FHNlL8MW9z3ff3dvnPPyH+y9fPLObS6RTmHviQrdu2smvXLubOnYuqqobjm2No07flFRXyAU7vOou9iw0xl04xfPjwqyy5vqc6vE1GRilYWiAsTPEJ90RvakRSfineLVzQm8lILCrBz98RLw87UivKMJiAl6stBmMBJgKMDWSqi8hQ52JpqsHatAY70xoibOywNVahN8TTzMIZe4USB0U19jIN9gpjLJx33PSxvBuuF7/vXvmR2H3nmbPvAy4cTcTWyYaAlr6oVWrGPvgwI8eOxMvKj4CWPmRezMHF14mKYiVyBxvMzY1J2n+eVZ+vp8PgNqxetZrRY0dzYksMEY/15mJSAV98+Qhmpka81O1dbJt7MeWDMaSnFRLo70TS8ST8OzUjJbeULq0DMVLIOXfoIkbGCoLb/3eG771V14ufwSA4fjQJHz8nipPzmD1pAVNnP07LXiEMHTqUX3/9lcUv/kxpXhmf73oPAK1Wz5FDl4ho6cWZ7bEsfXslQ6b2R+ajJjXJjMTEHEjMxMnajOfmPUGFQs57CzYTHeXPgUMXaaaVkakyMP6p7vTvGlrXMFWn1bHu6y2sW7iNTo/35oFHOpN6OhWfUC/yUmp7L/ny6UV4t/DgzZ9fvMNH8e64PHZWVla82OVt/MN9eGnRVACqK1Wc2HqGdgOi6p7OAMTExDCm3yM8EDWGafMn4d2i/rDmq389yrYtcXw8ayylGQVo1VpiUs9wYLeSLt1aMGp0e45vjqFtv5ZY2lqSm5LPty8spSS/tLZRlL0V7/z2ct3yTmyN4esXl9HnucGMm9wLE1Mjks6koiytQghRd12o26+SSpbM2sjJ87l46mooKKxgypdP0KlLCwwGA8c2ncajuTvpuRUE+jmRdCKJDoNbY2Zx7a4Uzx28wOxJC5g869F/ZXSsm3G1c+/whhMsenUFr/zwNOFdrl5l64svvsDa2pqnnnqqQev5/fff6+57mxbt4PevNvH+H9PrGpleTWWpkiWzNnI6tYQPPh2Nr58zH4yejapSVe8p8vWcO3gBY1NjWrQLqrsuBPo7knQ8ifYDozC3+vs7KYTghei38A315oUFk1k79082LdpBp2HtObLhOF7NPSjIKOLT7e/g9P8NYUsLynmt7wd0H9WZR94e0aBtup1u5gmxwWDgwQcfZNasWQQHX7v63uUuj19TlHgqmZoqNS27haLX6zm68RQ+oV6kxKbTuk/Ebe8v+j9fh/gvYdHB2LnYIpPL2LZtG59//jkA5pZmdRW2r6Z179pGCefS4u7Idt6rFi9ezOTJV28R+m8I7tAMU3MTjIwVnNgSg52rHRsXbMO/vTfh7UN5ZGJtl2uHDiRw4lIhXin5lOSUMHnWo8hkMtxdrLl4Ionk2DQ8AtwI7tAMBARGeGHjaI2piTFyhYxWvSJIPJVM6rGEunqBnr7OAPhc1tr8evWH/+nnmWsB6jqCvx/J5TI6dq798SCr0dCstT9bl+zmTPJJpk+fjrOzM5Hdw9i36jA/f/YHuW42RAR50L9H7Y3cs5kbHYe0pde4aI6cPkxEpDeFpy4hszSh57hoLGzMWT59Bc087AkJcMMgINzDmb0/HeDir4fp1b4ZciMFi77dRUCQKz3GdqYwq5jeA1vi5e+K1//XD/cN8WLVL0exigyiZXv/u3a87iaZTEa7Aa3qDWJkYW1+1evi6tWreWnSK/g386tLhtPjM1n9xUaGvziYwGautCsJxMLSBJf/b7SUkH+RyHBbLqw/ylk3a7qP6sTZAxfYvnwvj84YRc+Hu7Bt2R4Mej1RPcJqk9o3VtJxaFvcA13p2D+SPgMiMDGtvV0FRV07TjYO1vR9pBum+y4QYGvC3p8PEPvHEc7/eQIHd3uGvzAYAN//33ZPP+cbHh9HTwfaDWhVr5eGLT/sIu1cJpNnjb/qqGd3g3uAK+0HtsLR8+q9n9TU1JCens7XX39db/r+NUc4s/sckz4bX+/HD4Cx8d/VUXxDvWg3oBXWDtdPQnKS8ihMyCI8wh8bm9rlRfUIR12tbvC+hHcJoaZazfznfyC0cwt6jetCcmwaJ7bE4BfuXa9niL++vy4+TlQUV5J0Jo2ASD+COwQhDAaatwskNTYdC2tzfvnkD/Q6PSNeHkK7Aa0Iat10zvmTJ08ydOjQBifDUD9+TdHl1SAUCgXRD9Z2fOB9nR9kt+LeOJPvgN6PdEUIwWdvz2bSyKmYmZmhqqph/+ojtOodgYu3EznJeZw/lED3MZ3r+jH8y+LFixk69O/K3Lmp+Zzdf4EeYzpjYmZyp3fnjps5c+YdTYi7j+pE91GdEEKQcjYD57Iq4mMucj7pHB8s/vspQ0FBBSnJBQi05KXk1XXqbeNgzfBpg/jiyQWkiUv0HBNNz8sawpyJSUej1jL61WG8+8Bn5KVeOcx32vlMUmLT6D66c73RcG4kPT7z1nb+P8Y9wJV+E3rywTOf0tunO90mdQdg8JS+nNwey8kdsRS38sXR9u83NM3bBNZdDBc/vZiPX5uFrrAUuVyOdwsPVMoachNyeKBPJH07h9CvcwiZCdkcrqwiPaUcvU7Psa1niDuVipm5MU6DInl67oQrtu1seh5HYpKxdXNgyD+6VSwrLOfoxlN0Gd7hqm+P/itkMhmPvjvqhvOtXbuWmTNnXtHpfUWxksRTyexYsY/xb4+gdZu/k4xjm04xb87XrPplFR+uOUhhZhEAJbmltY1VK1R4NnMHg6BGqcEnxAuNSkPK2XQCIv3oOKQNz371xBXbcmpHLDK5nNa9I1Cr1OxbdYSW3UNx83PB0dqEQGtjuo/sSNLJZFLPZaJQyEk6k0Z4l+BG1zV093e9ov/5nKQ8UuLSMBjunRes/uE+1+wnX6VSsXv3bubNm3fFZ/lphaScTUer1sI/EuLL73vhXUKu+eT5cmUF5RQk59Eswgdzs9r76F9tBxoibn88KmUN4dEtSIlLx8HdnqryKnb+tJ/EU8lUFldeUeav7292Ui7ZibkEtvKjZbfQv6/5j9T+L/1CJjqNDhNTYwIjfXFrwA+ie8G5c+ewsbFp9D34n3mL5AZuW+WLW3Tu3DkRGhraqDKNrYfzzTffiFZ2ncTTbacLIYQ4e/CCGGAyRvzyye9CCCEWTV8hBpqNEyln00VlqVKU5JXWlf1nXZwlb60UA03HikunU+pNn9L+LTHA7RkxIPAV0T/iLdG3w/uiV4+PRbeBn4nOIz4X7cfPFq0nzRGRz80R4f+bI0LemiOafzBHBM6aLfy+/kL4Lpol+q5fLAZsWShC180Q7ba8IXrtfkkMP/S0mHziMfFa7AjxeXw/sSyxk9iUEiaOp/uK1Ew3UVXQp24blEqlGDFihOjYsaPo2LGjOHr0qBg8eLDo1q2b6Natmzhz5owQQghfX9+6Mt27dxepqalixowZYvz48aJ///4iIiJCnDhxQnzyySfC1NRUvPzyyw2OTUM0NH5FOSWiskwp+vceIKKNB4il7/wiDAaDyEzMqZunqqK6Xl3Gy12tHtVTTywW40bW1jHOSysQapX6ink+n/iNGGL1iCjMKmrMbt03Ghq/rEs5Ijc3V4QHtRR9jUeIuP3xdZ8VZheLB+0eEy92fateDDIz/j7mQ4cOFS90eUs85Pi4eNBhgni5x7tXXc9XzywWgy0eFjnJeaK8uEKMcJl4RR1TIWrrUWZdqv3uPPXNWtHltW+E5ir1hrcs2S36GY0WB/84ds19y0nJq1d/sqlo7LVz8+bNYtq0adf8fPl7v4l+RqPF0U2n6k0f4zlZ+NkFXXfZbw6eKR60f1w84ve0eDL8RVFdWV33WWZC9lXrsE5o8XxdHdOEk0lioOlYsezdX2u3ZcZvYqDpWHHxRFLd/FUV1eIBu8fEOw98euOd/X95aQVCXaNp8Px3UmPjN23aNLF58+ZGr+dm66Cu+GC1GGAyRsQfTWx02WfbvybGek2pN+3onydFP6PR4sf3V92w/LZlu0VfxSjxx7zNQqfTieyk3CvmyU8vEEMsHxFfPPlto7fvdmhM/Kqrq0Xv3r1Ffn5+o9fT1OsQN4S6RiNyUxt/bK7mnkiIly9fLry8vOolaA3RmC+VRqMRTz/9tFBV19RrsHH5xffyv58Me1EMtR5fNz07O/uKZf6zrBD3RkL8+eefixkzZgghhDh48KAIDAwUixYtEkIIcf78edGuXTshxLUT4hdeeEEIIcSKFSvEU089dcW8t0tDG9UNNB0rRoU8Lr7++uu6Y/7LJ7+LPrKRdcnKM+1eEwNMxgid7srE5mqx02p1Qq3WisKsItFPMUrMeOizq66/Wqm6mV27LzQkfie3nxF9ZCPFI/0fF7GxsVecM8OdJoqHfZ8S/RSjxNtDPxFCCLH+j5Oid9ePxLbNsUKI2vjpdDox2n2SGOf91FVj/Je/lv+w71Qx0vXJq867+LUVoo9spIjZfVbo9QahUl87ob3aOf6XtPMZoq98lJg1cf4157lXNTaheu6550R19dWPxdaltQ3l1szdeMVnapVaZKRnXHfZer1eqKpqhLKiWgwwGSOe7/SGEEKIfasPiz6ykWLV5+uvKKNRa+r9ELnWdfxyqqqaGzYQ+0tB5v9fF4bPatD8d1pj4peXlyfefPPKH4YNcbVrZ0Nd79y5Hp1Wd9UHFGvmbBB9ZCPFtmV7rln23KGLoo9spJj33GIhhBALXloq+shGirMHL9y27bsdGhO/tWvXik2bNt3Uem4lfk3FR2PniL7yUSInOe+Wl3XXq0yUlJSwYcMGfv75Zx577LF/ZR0ajYbJ/Z9DnmuGQiGv9/r78sr5l//9/rrpdSP8AMyfP5+PP/74qvNeLv1CNjJLi1va3sSyIozMdJheu53HFYQuse7fSUlJPPTQQwBER0fTvHlzunbtCkBoaCj5+flXlr+sbWVkZG0n515eXtTU1NzMLtw2MpmMF1Y+iYmFET0H9Kib/uC0QTRvG0hkj9q+Yt9Z9TKVJUoUiiurNlweuwUvLePw+uN8c+IzzK3NeLX3+7TpF8kLC6Zcdf232q/h/S6qVzgPfNybMRNG4Ox25evJL3bPwNjUmPKiStz8a+uwDhgYiaenPS2jahtL/RW/WTvfBZnsqjH+i7mVOR+NnYNcIcfS1pwPRs7m/T+m15tn3JvDad03kpbdQ5HLZZhdZ4CFq53jf/EN9eblxVP54c2fWTd/Cw8+N/C6x6IpEkLwxx9/XPGqvaiokuenLqNv/3AefbwLTp4ORHQLvaK8iZkJCxYu4OOPPybhZDLvDP2ECR+MZdDkPnXzyOXyukZtXx6aWddhf/SD7fls+zuERbdg9eyNrJmzkc93zcAn2POKrtaudR2/3PUazv2Ts5cjs/d9UG/UrqaosLCQhIQEZs6ceVPlr3bfa6jrnTvXozBSXLWK2tCn++MX7kPL7ld+z/4S0rEZs3a+S/M2tSONPjpjFB2GtCG0U/N68xXnlvJ8xzfo/XBXnvzkkZvazjvh8OHDdOrUCXf3hg15/E+3Er+m4pmvnmDYMwPq1fG/WXc9IXZwcGDNmjWkpaXdcF61Wo1a/XeSWlFR0aB1vPPOO7ToGESwa9g164Ke3hnHhgXbmDLrUTwC3fAMcsdgMDD7yW/xauHJuXPn6uY9s+cc6+ZvYdKn4xs8NvedFBwczIkTJ+jfvz9HjhyhRYsWHDp0iODgYOLj47G1re3HUaVSoVKp0Ol0JCUl1ZX/Zx1BqJ8w36ybiZ9Go2H2wlksXbq03nQzC1Na92lZ97ebn0u9RkGXuzx2HYe2wcHdHnNrMxb+bzmuvi4Mfbo/9q52jdyb+8+14ndg7VFitp3j+fmTsHWq32L60qVLbD+8mWemX/0Hx18DWXhddr8yMTWiTbu/h87+K36+oX8PnFOYVcyCl5bS97EedBr6dx+mAN1HRxMeHYJcIcfW6cqhdS1tLOoaywIse+dXqiqqr1pX9UY6P9iOwqxiwrs0vKHL3XCz187Xn3uLMwfP0rV993rDuFpbmzH0wdZERHhjbGJMm76R11zGX/Fz83Nm6NT+hFyWnGxduofTO2N5YcEULG0saHHZqHgKI0XdOd6yeyjqajWO7nYN2u7boTENaf9tNxM/IQTPPvssb7755k2v9/Jr591mYmZy3e8Z1P64+quh9MqZa3n4rRG06hXBxoXbOX/4Ii999xSm5qZY2Vkw7On+hHZucd3l6XV65kxZSECE7y0N5nEz8SssLOSdd95h48aNN73eeyl+/xZ7F1vsXW5PX+J3PSFujE8++YT333+/UWVKSkrQ6/W8+dn0a85TkFnEnl8PUVmsRK/T1/usqkJF3P54rMz+vtFr1VpUlSr0Wl3jduAOmTJlChMnTqRHjx4IIZg1axYzZ85k+fLlaDQaFi+u7Rz9mWeeoVOnTvj7+xMYeP2GJgEBAUydOpWFCxfe9HbdTPx27tzJU089haen541nvgZfX18yE7I5vjmGAU/0rBvOVKWswS/c+4qEKutSLkc3nqT/xJ63vVuXpuxa8dOotVRX1tQN6Xu5lStX8u23395SR/i+vleO/qbX6akoVnJgzVH8wrzrPR3oOrxDo5avUtagqry5NyE2DtYNapB2t93MuafT6biQeIFW7u2uiK2pqfFVRyW8mr/iZ+tkw6Mz6h8rjUqDSlkDN/jB3aJtYL1k+d8khGDrkt24+btcc+jjO+1m4nf27Fk6d+5MVFTUTa/3audeU6DT6qmurKkbHU9drUZVWVP3NTM1N2Xs6w/dcDlCCFSVKmoa0UPG1dxM/H799Vc+//xzLCxu/o1zU43fXXPLlS5uwjvvvCMiIyNFZGSkWL++tn5YamrqDeup1tTUiPLy8rr/MjMzr1sPp7S0VOzevfuG27Nr5QExxOoRcfTPk1d8Vl5cIUa5PSk+enT2jXdMcl2Njd+xY8dEZmbmLa+3oqJCrJ37pxhm82i9hjbX8se8zWKYzaPi/OGLt7zu/5LGxm/dunXXre/bUBUVV3beL4QQZ/acE0Otx4tNi3bc8jr+6xobO7VaLTZs2HBb1n2t+N2r1Cq1GOs1Rbw/6ou7vSl1Ghu/xMREERcXd8vrbWqx+6ea6poGXfP/9e1oZPy2b98uKitvfRCRph6/O+2eaFQnRMMS4n+6XsV0g8Egxo8fL44cOdKgZV3vxq3T6e6L1pp32vXiV1paKnr27HlbTui/YteY5KzmKo06JPVdL3779u0TTzzxxG1Zz/XOvduRcN+PbtSo58033xSrVt24RX9DNMVr573+vbpe/DQajejbt6/IyLh+Y8aGaIqxu9w3LywRfWQj6/Vscy+4XvwSExPFkCFDbst3sKnH705rUlUm/kn8//uPq9XHuXTpEgEBAYSGhja4vtz1aLXa27Kc/wJra+ur1jNurOvFb9myZbz55psIIW75uDc2dhsXbmfFB6t5/4/phHT4741Mdyfit3LlSt5//33p3LvN7kTsSktLKS0tpX///lL8brM7Eb8dO3Ywbtw4bG1t7/i1814z+q1hdBrRBt9Iz9uyH3cift9//z2fffYZVVVVt7yeph6/2+1G8WvSQzdnZWXh7e194xklt1Vjhpu8Hil+d4cUv6ZLil3TJsWvaZPi17TdKH5NOiE2GAzk5OTcMOuvqKjA29ubzMzMm/oy3+/l/+l2/UpuSPzu9r439fJXI8Wv6ZT/pzsZO7j7+9/Uy/+TdO41vW24nBS/plP+am4UvyZdZUIul+Pl5XXjGf+fjY3NLR3Y+7387daY+N3tfW/q5f8NUvzuXPnbTbp23tnyt9v9dO7dK9twO91P8bvb5Rvj5vtCkkgkEolEIpFI/gOkhFgikUgkEolEcl+7LxJiU1NTZsyYgWljxkKWyt8T7va+N/Xyd9vd3v+mXv5uu9v739TL3013e99vx7G7F7bhbrnb+97Uy9+MJt2oTiKRSCQSiUQiuVX3xRNiiUQikUgkEonkWqSEWCKRSCQSiURyX5MSYolEIpFIJBLJfe2+SYjPnz9PWFhYo8utXLmS0NBQmjVrxvz5829q3ZWVlURERJCWltbosnPmzCEsLIyIiAieeOIJNBpNo8q/8cYbhISEEBoaypw5cxq9/nvF3YrfrcQOpPiBdO415diBFD8pflL87ibp3ncH4yfuA8uXLxdeXl7C19e3UeWysrKEj4+PKCoqEkqlUrRs2VLExcU1ahlHjx4VkZGRwtjYWKSmpjaq7LFjx0R4eLhQKpXCYDCI8ePHizlz5jS4/KZNm0TXrl2FVqsVVVVVws/PT1y8eLFR23AvuFvxu5XYCSHFTwjp3GvKsRNCip8UPyl+d5N077uz8fvPPyEuKSlhw4YN/Pzzz40uu3PnTnr37o2joyOWlpaMHDmSNWvWNGoZCxcuZN68eXh4eDR6/fb29syfPx9LS0tkMhmRkZFkZGQ0uPygQYPYtWsXRkZGFBYWotfrsbS0bPR23E13M363EjuQ4iede003diDFT4qfFL+7Sbr33fn4/ecTYgcHB9asWYOPj0+jy+bk5NT7Qri7u5Obm9uoZSxdupSuXbs2et0AzZo1o3v37gAUFBQwf/58hg0b1qhlGBsb8/bbbxMSEkKvXr3w9PS8qW25W+5m/G4ldiDFTzr3mm7sQIqfFD8pfneTdO+78/H7zyfEt8JgMCCTyer+FkIgl9/5Q5aWlkbPnj2ZPHkyPXv2bHT5jz76iKKiIrKysli8ePG/sIX3Jil+TZcUu6ZNil/TJsWvaZPid3P+kwnxu+++S1RUFFFRUWzYsOGml+Pl5VXvV1VeXt5Nv0K4WWfOnCE6OpqpU6fy1ltvNars+fPnOXv2LAAWFhY89NBDxMXF/RubeVtJ8avVFOMnxa5WU4wdSPH7ixQ/KX53gxS/Wnctfv96LeV7RGpq6k1VTPf19RX5+flCqVSKiIgIcezYsZtav6+vb6MrpxcUFAhXV1exdu3am1rn6tWrRefOnYVarRY1NTWiV69e4rfffrupZd1tdzN+NxM7IaT4/UU695pu7ISQ4ifFT4rf3STd++5c/P6TT4hvF09PTz7++GN69uxJq1atGD9+PO3bt79j6//yyy+pqKjggw8+qPvV2JhfWiNHjqR79+5ERUXRtm1bevTowejRo//FLb63SPFruqTYNW1S/Jo2KX5NmxS/myMTQoh/fS0SiUQikUgkEsk9SnpCLJFIJBKJRCK5r0kJsUQikUgkEonkviYlxBKJRCKRSCSS+5qUEEskEolEIpFI7mtSQiyRSCQSiUQiua9JCbFEIpFIJBKJ5L4mJcQSiUQikUgkkvualBBLJBKJRCKRSO5rUkIskUgkEolEIrmvSQmxRCKRSCQSieS+JiXEEolEIpFIJJL7mpQQSyQSiUQikUjua1JCLJFIJLfB3r17cXNzo0ePHnX/bdiw4Zrz+/n5XTHtvffeY9myZZw5c4ZPP/30lrZn0aJFODs7U11dXW/6smXLiI6Oplu3bkRHRzN37txbWo9EIpH8Fxjd7Q2QSCSS/4oBAwawbNmyW15OVFQUUVFRt7SMH3/8kUcffZRffvmFJ598EoD169ezZs0atm/fjqWlJTU1NTzwwAP06dOHiIiIW95uiUQiaaqkJ8QSiUTyLyorK2PIkCF0796d7t27ExsbW+/zffv20aZNG/r06cO+ffuA2qfNEyZMIC0tjc6dOzNy5EgiIyN5//33Afjjjz9o1aoVvXv3ZsCAAezdu7feMi9cuICZmRnPPfcc3333Xd30BQsWMHv2bCwtLQEwMzNj69atUjIskUjue1JCLJFIJLfJ1q1b66pLjBo1CoCPPvqIBx54gH379rFgwQImT55cr8xrr73G77//zs6dO7Gzs7timampqaxYsYJjx44xb948AN544w12797Njh07qKmpuaLMDz/8wKOPPkpAQADm5uacOnUKgLS0NAICAgBYs2YNPXr0oF27dnz55Ze38ShIJBJJ0yMlxBKJRHKb/PW0du/evaxevRqA+Ph4unbtCkBoaCj5+fn1yhQUFODr6wtAdHT0Fcts3rw55ubmmJmZYWVlRWFhIfb29tjb2yOXy+nYsWO9+bVaLb///jvDhw8HYMKECSxcuBAALy8vUlJSABg5ciR79+7lueeeo6ys7PYdBIlEImmCpIRYIpFI/kXBwcEcOnQIqE2ObW1t631ua2tLcnIyQN2T3MvJZLJ6fzs6OlJSUkJZWRlCCI4fP17v8z///JPOnTtjbW0NwKhRo9i4cSMVFRU8++yzvPrqq3UN7bRaLUePHr1iHRKJRHK/kRrVSSQSyb/orbfeYuLEiSxfvhyNRsPixYvrfb58+XLGjh2LjY0NpqamN1yeXC7niy++oG/fvtja2qJSqTAy+vtSvmTJEmJjY2nbtm3dNKVSyYoVK3j22Weprq5m4MCByOVyysvLGTx4MP/73/9u3w5LJBJJEyQTQoi7vRESiUQiabjPPvuM//3vfxgZGdGzZ0+WLl161W7cJBKJRNIw0hNiiUQiaWJMTExo27YtlpaW9O3bV0qGJRKJ5BZJT4glEolEIpFIJPc1qVGdRCKRSCQSieS+JiXEEolEIpFIJJL7mpQQSyQSiUQikUjua1JCLJFIJBKJRCK5r0kJsUQikUgkEonkviYlxBKJRCKRSCSS+5qUEEskEolEIpFI7mtSQiyRSCQSiUQiua/9H7CLbzZ/k9eqAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 700x170 with 7 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.linear_model import LinearRegression\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.stats import gaussian_kde\n",
    "from scipy.stats import pearsonr\n",
    "import math\n",
    "\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "\n",
    "xlim=(min(predicted_stability_vs_binding['folding_ddG']),max(predicted_stability_vs_binding['folding_ddG']))\n",
    "ylim=(min(predicted_stability_vs_binding['binding_ddG']),max(predicted_stability_vs_binding['binding_ddG']))\n",
    "\n",
    "fig, ax = plt.subplots(nrows=1, ncols=6, figsize=(7,1.7))\n",
    "coliterator=0\n",
    "rowiterator=0\n",
    "for hamm_dist in range(2,8,1):\n",
    "    y = predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist][['binding_ddG']].to_numpy()\n",
    "    x = predicted_stability_vs_binding.loc[predicted_stability_vs_binding['Nham_aa_x']==hamm_dist][['folding_ddG']].to_numpy()\n",
    "    \n",
    "    model = LinearRegression()\n",
    "    model.fit(x,y)\n",
    "    \n",
    "    y_pred = model.predict(x)\n",
    "    \n",
    "    ax[coliterator].set(xlim=xlim,ylim=ylim)\n",
    "    hbin=ax[coliterator].hexbin(x,y,gridsize=50,bins='log', vmax=50, rasterized=True,linewidths=0, extent=[min(predicted_stability_vs_binding['folding_ddG']),max(predicted_stability_vs_binding['folding_ddG']),min(predicted_stability_vs_binding['binding_ddG']),max(predicted_stability_vs_binding['binding_ddG'])])\n",
    "    ax[coliterator].set_title(str(hamm_dist)+'\\nr = '+str(round(pearsonr(x.ravel(),y.ravel())[0],2)), fontsize=7)\n",
    "    ax[coliterator].axline((0, 0), slope=1,ls='dashed',color='k',lw=0.5)\n",
    "    ax[coliterator].set_ylim(min(predicted_stability_vs_binding['binding_ddG']),max(predicted_stability_vs_binding['binding_ddG']))\n",
    "    ax[coliterator].set_xlim(min(predicted_stability_vs_binding['folding_ddG']),max(predicted_stability_vs_binding['folding_ddG']))\n",
    "    ax[coliterator].spines[['right','top']].set_visible(False)\n",
    "    ax[coliterator].set_xticks(ticks=np.arange(-1,4,1), labels=np.arange(-1,4,1))\n",
    "    ax[coliterator].set_yticks(ticks=np.arange(-1,4,1), labels=np.arange(-1,4,1))\n",
    "    #ax[coliterator].set_xticklabels([0,-0.5,-1])\n",
    "    ax[coliterator].tick_params(axis='x', labelsize=8)\n",
    "    ax[coliterator].tick_params(axis='y', labelsize=8)\n",
    "    ax[coliterator].axhline(0, ls='--', color='k',lw=0.5)\n",
    "    ax[coliterator].axvline(0, ls='--', color='k',lw=0.5)\n",
    "    ax[coliterator].set_rasterization_zorder(0)\n",
    "    if coliterator != 0:\n",
    "        ax[coliterator].yaxis.set_ticklabels([])\n",
    "    coliterator+=1\n",
    "    \n",
    "#fmt = '%1.f'\n",
    "#cax=fig.add_axes([0.99,0.25,0.005,0.3])\n",
    "#cb=fig.colorbar(hbin, shrink=0.7,format=fmt,cax=cax)\n",
    "#cb.ax.set_title('Counts',fontsize=7)\n",
    "##cb.set_label('counts', rotation=270, fontsize=7, labelpad=8)\n",
    "#cb.set_ticks([1,5,10,50])\n",
    "#cb.set_ticklabels([1,5,10,50], fontsize=6)\n",
    "\n",
    "fmt = '%1.f'\n",
    "cax=fig.add_axes([0.1,0.22,0.045,0.05])\n",
    "cb=fig.colorbar(hbin, ax=ax, shrink=0.7,format=fmt,location='top',cax=cax)\n",
    "cb.outline.set_visible(False)\n",
    "cb.ax.tick_params(axis='both', which='both',direction='in', color='white',width=0.25)\n",
    "cb.ax.xaxis.set_tick_params(pad=0)\n",
    "#cb.ax.set_title('Count', fontsize=7)\n",
    "#cb.set_label('count', rotation=0, fontsize=7, labelpad=2)\n",
    "#cb.ax.xaxis.set_label_position('right')\n",
    "cb.set_ticks([1,10,50])\n",
    "cb.set_ticklabels([1,10,50], fontsize=6, rotation=0)\n",
    "ax1 = cb.ax\n",
    "ax1.text(60,0.1,'count',rotation=0, fontsize=7)\n",
    "\n",
    "fig.text(0.5, -0.015, 'Folding ΔG', ha='center', fontsize=7)\n",
    "fig.text(-0.005, 0.5, 'Binding ΔG', va='center', rotation='vertical', fontsize=7)\n",
    "fig.tight_layout()\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "#plt.savefig('FYN_binding_vs_stability_ddG_2ndordersprsreg3statemodel_byNham_correl.svg', dpi=300, transparent=True, bbox_inches='tight')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}