matrix_responder_timepoint_fourmer.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "\n",
    "from fragment_count.views import plot_motif\n",
    "from fragment_count.utils import entropy, pool, safe_normalise\n",
    "from source import load_fragment_statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "stats_per_response = load_fragment_statistics(statistic='fourmer')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "resp_normal, resp_var = stats_per_response['responder (pr+cr)']\n",
    "non_resp_normal, non_resp_var = stats_per_response['non responder (sd+pd)']\n",
    "# Non-evaluables belong to the group of non-responders.\n",
    "noneval_normal, noneval_var = stats_per_response['non evaluable (ne)']\n",
    "non_resp_normal = non_resp_normal.append(noneval_normal)\n",
    "non_resp_var = non_resp_var.append(noneval_var)\n",
    "\n",
    "# Override `stats_per_response` to only include response and non-response.\n",
    "del noneval_normal, noneval_var\n",
    "stats_per_response['non responder (sd+pd)'] = (non_resp_normal, non_resp_var)\n",
    "del stats_per_response['non evaluable (ne)']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Patient ID  sample number\n",
       "1023        0                49956\n",
       "            1                29757\n",
       "1031        0                 4257\n",
       "            1                 5095\n",
       "1032        0                49691\n",
       "                             ...  \n",
       "972         1                61929\n",
       "979         0                31569\n",
       "            1                22142\n",
       "982         0                25834\n",
       "            1                 4270\n",
       "Name: count, Length: 90, dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resp_normal.sum(axis=1).groupby(['Patient ID', 'sample number']).sum().rename('count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25615.088888888888 +/- 27421.93414598279\n",
      "1399.3 +/- 3683.2631460935277\n"
     ]
    }
   ],
   "source": [
    "def count_4mer(data_frame):\n",
    "    return data_frame.sum(axis=1).groupby(['Patient ID', 'sample number']).sum().rename('count')\n",
    "\n",
    "print(count_4mer(resp_normal).mean(), '+/-', count_4mer(resp_normal).std())\n",
    "print(count_4mer(resp_var).mean(), '+/-', count_4mer(resp_var).std())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7fce37169cd0>"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.catplot(\n",
    "    data=resp_normal.sum(axis=1).groupby(['Patient ID', 'sample number']).sum().rename('count').reset_index(), \n",
    "    x='sample number', \n",
    "    y='count',\n",
    ")\n",
    "\n",
    "sns.catplot(\n",
    "    data=non_resp_normal.sum(axis=1).groupby(['Patient ID', 'sample number']).sum().rename('count').reset_index(), \n",
    "    x='sample number', \n",
    "    y='count',\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "motifs = ['CCCA', 'CCAG', 'CCTG', 'TAAA', 'AAAA', 'TTTT']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x216 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rn0 = resp_normal.loc[:, \"0\",:].sum(axis=1).groupby('Patient ID').apply(safe_normalise)\n",
    "rv0 = resp_var.loc[:,\"0\",:].sum(axis=1).groupby('Patient ID').apply(safe_normalise)\n",
    "plot_motif(\n",
    "    motifs,\n",
    "    rn0.groupby('4mer').mean(),\n",
    "    rv0.groupby('4mer').mean(),\n",
    "    rn0.groupby('4mer').std(),\n",
    "    rv0.groupby('4mer').std(),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalise_per_patient(data_frame, time_point):\n",
    "    return data_frame.loc[:, time_point, :].sum(axis=1) \\\n",
    "        .groupby('Patient ID') \\\n",
    "        .apply(safe_normalise)\n",
    "\n",
    "def time_pair_plot(clinical_category, plot_i, motifs):\n",
    "    normal, var = stats_per_response[clinical_category]\n",
    "        \n",
    "    p0_normal = normalise_per_patient(normal, time_point=\"0\")\n",
    "    p0_var = normalise_per_patient(var, time_point=\"0\")\n",
    "    \n",
    "    p1_normal = normalise_per_patient(normal, time_point=\"1\")\n",
    "    p1_var = normalise_per_patient(var, time_point=\"1\")\n",
    "\n",
    "    # Responder, baseline.\n",
    "    ax = plt.subplot(3, 2, plot_i)\n",
    "    ax.set_title('Baseline')\n",
    "    plot_motif(\n",
    "        motifs, \n",
    "        p0_normal.groupby('4mer').mean(), \n",
    "        p0_var.groupby('4mer').mean(), \n",
    "        p0_normal.groupby('4mer').std(), \n",
    "        p0_var.groupby('4mer').std(),\n",
    "        ax=ax,\n",
    "    )\n",
    "    ax.set_ylabel(clinical_category)\n",
    "    ax.set_ylim([0.0, 0.013])\n",
    "\n",
    "    # Responder, follow-up\n",
    "    ax = plt.subplot(3, 2, plot_i + 1)\n",
    "    ax.set_title('Follow-up')\n",
    "    plot_motif(\n",
    "        motifs, \n",
    "        p1_normal.groupby('4mer').mean(), \n",
    "        p1_var.groupby('4mer').mean(), \n",
    "        p1_normal.groupby('4mer').std(),\n",
    "        p1_var.groupby('4mer').std(),\n",
    "        ax=ax,\n",
    "    )\n",
    "    ax.set_ylim([0.0, 0.013])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,8))\n",
    "\n",
    "for i, response_category in enumerate(stats_per_response.keys()):\n",
    "    time_pair_plot(response_category, 2*i + 1, motifs)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('figs/motifs.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Entropy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def entropy_per_patient(data_frame, time_point, drop_zero=False):\n",
    "    S = data_frame.loc[:, time_point, :].sum(axis=1) \\\n",
    "        .groupby('Patient ID') \\\n",
    "        .apply(safe_normalise) \\\n",
    "        .groupby('Patient ID') \\\n",
    "        .apply(entropy)\n",
    "    if drop_zero:\n",
    "        return S[S != 0.0]\n",
    "    return S"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.91270257 0.74145361 0.90395291 0.64158417]\n",
      "[0.03943303 0.20129087 0.04368223 0.2190695 ]\n",
      "[0.90487491 0.74065959 0.91070103 0.76442805]\n",
      "[0.05263055 0.19232852 0.05301272 0.16719498]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,3))\n",
    "\n",
    "# Reponders\n",
    "rn0 = entropy_per_patient(resp_normal, time_point=\"0\", drop_zero=True)\n",
    "rv0 = entropy_per_patient(resp_var, time_point=\"0\", drop_zero=True)\n",
    "\n",
    "rn1 = entropy_per_patient(resp_normal, time_point=\"1\", drop_zero=True)\n",
    "rv1 = entropy_per_patient(resp_var, time_point=\"1\", drop_zero=True)\n",
    "\n",
    "plt.subplot(1,2,1)\n",
    "plt.title('Responders')\n",
    "plt.rc('font', family='serif')\n",
    "labels = ['wild type ($t_0$)', 'variant ($t_0$)', 'wild type ($t_1$)', 'variant ($t_1$)']\n",
    "items = [rn0, rv0, rn1, rv1]\n",
    "\n",
    "mean = np.array([i.mean() for i in items])\n",
    "std = np.array([i.std() for i in items])\n",
    "print(mean)\n",
    "print(std)\n",
    "\n",
    "plt.errorbar(x=labels, y=mean, yerr=std, fmt='o')\n",
    "plt.ylabel('Entropy')\n",
    "plt.ylim([0.5, 1])\n",
    "plt.xticks(rotation=45)\n",
    "\n",
    "\n",
    "# non-reponders\n",
    "nrn0 = entropy_per_patient(non_resp_normal, time_point=\"0\", drop_zero=True)\n",
    "nrv0 = entropy_per_patient(non_resp_var, time_point=\"0\", drop_zero=True)\n",
    "\n",
    "nrn1 = entropy_per_patient(non_resp_normal, time_point=\"1\", drop_zero=True)\n",
    "nrv1 = entropy_per_patient(non_resp_var, time_point=\"1\", drop_zero=True)\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.title('Non-responders')\n",
    "plt.rc('font', family='serif')\n",
    "labels = ['wild type ($t_0$)', 'variant ($t_0$)', 'wild type ($t_1$)', 'variant ($t_1$)']\n",
    "items = [nrn0, nrv0, nrn1, nrv1]\n",
    "\n",
    "mean = np.array([i.mean() for i in items])\n",
    "std = np.array([i.std() for i in items])\n",
    "print(mean)\n",
    "print(std)\n",
    "\n",
    "plt.errorbar(x=labels, y=mean, yerr=std, fmt='o')\n",
    "plt.ylabel('Entropy')\n",
    "plt.ylim([0.5, 1])\n",
    "plt.xticks(rotation=45)\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.savefig('figs/entropy_fourmer_response.png')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}