notebooks/NovelCandidates.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Novel BGC candidates overview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 822,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "from matplotlib import pyplot as plt \n",
    "from IPython.core.display import display, HTML\n",
    "import seaborn as sns\n",
    "import sys\n",
    "sys.path.append('../bgc_detection/candidates')\n",
    "from candidate_regions import candidate_regions\n",
    "from biosynthetic_pfams import AS_BIO_PFAM_IDS\n",
    "from ipywidgets import interact\n",
    "from Bio import SeqIO\n",
    "import matplotlib\n",
    "font = {'family' : 'Arial', 'size'   : 12}\n",
    "matplotlib.rc('font', **font)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 673,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import sys\n",
    "\n",
    "sys.path.append('../bgc_detection/similarity')\n",
    "import group_scatter_plot\n",
    "\n",
    "from group_scatter_plot import group_scatter_plot, apply_group_color"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 674,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>clan_id</th>\n",
       "      <th>description</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>pfam_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PF00001</th>\n",
       "      <td>CL0192</td>\n",
       "      <td>7 transmembrane receptor (rhodopsin family)</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        clan_id                                  description\n",
       "pfam_id                                                     \n",
       "PF00001  CL0192  7 transmembrane receptor (rhodopsin family)"
      ]
     },
     "execution_count": 674,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pfams = pd.read_csv('../data/pfam/Pfam-A.clans.csv').set_index('pfam_id')\n",
    "pfams.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 675,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "DEEPBGC_PATH = '128lstm-100pfamdim-8pfamiter-posweighted-neg-10k-fpr2'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 947,
   "metadata": {},
   "outputs": [],
   "source": [
    "paths = [\n",
    "    '{}/candidates.csv'.format(DEEPBGC_PATH),\n",
    "    'antismash-minimal/candidates.csv',\n",
    "    'antismash-clusterfinder/candidates.csv',\n",
    "    #'antismash-clusterfinder-default/candidates.csv',\n",
    "    #'antismash-export/candidates.csv',\n",
    "    #'hmm_clusterfinder-tpr80-postproc/candidates.csv'\n",
    "]\n",
    "labels = [\n",
    "    'DeepBGC',\n",
    "    'antiSMASH',\n",
    "    'ClusterFinder',\n",
    "    #'antiSMASH+ClusterFinder',\n",
    "    #'antiSMASH export'\n",
    "    #'HMM Clusterfinder tpr=80%'\n",
    "]\n",
    "MODEL_COLORS = {\n",
    "    'DeepBGC': 'red',\n",
    "    'antiSMASH': 'orange',\n",
    "    'ClusterFinder': 'dodgerblue'\n",
    "}\n",
    "colors = [MODEL_COLORS[l] for l in labels]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 922,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "BIO_COLORS = {\n",
    "        'Polyketide' : 'deeppink',\n",
    "        'NRP' : 'lime',\n",
    "        'RiPP' : 'blue',\n",
    "        'Other' : 'darkcyan',\n",
    "        'Saccharide' : 'orange',\n",
    "        'Terpene': 'turquoise',\n",
    "        'Alkaloid' : 'purple',\n",
    "        'No class' : 'grey',\n",
    "    'No confident class' : 'grey',\n",
    "        'Hybrids' : 'yellow',\n",
    "}\n",
    "ACT_COLORS = {\n",
    "    'Antibacterial': 'deeppink',\n",
    "    'Antifungal': 'orange',\n",
    "    'Cytotoxic': 'blue',\n",
    "    'Inhibitor': 'darkcyan',\n",
    "    'No class' : 'grey',\n",
    "    'No confident class' : 'grey',\n",
    "    'Hybrids' : 'yellow'\n",
    "}\n",
    "SPECIES = [\n",
    "    'Mycobacterium','Bacillus','Pseudomonas','Streptomyces','Salmonella',\n",
    "    'Escherichia','Paenibacillus','Amycolatopsis','Staphylococcus','Clostridium',\n",
    "    'Xanthomonas','Klebsiella','Frankia','Nocardia'\n",
    "]\n",
    "SPECIES_COLORS = {s: matplotlib.cm.tab20.colors[i] for i, s in enumerate(SPECIES)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 679,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "model\n",
       "ClusterFinder    62491\n",
       "DeepBGC          10926\n",
       "antiSMASH        13865\n",
       "dtype: int64"
      ]
     },
     "execution_count": 679,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "candidates = []\n",
    "for path, label in zip(paths, labels):\n",
    "    cands = pd.read_csv('../data/bacteria/candidates/{}'.format(path))\n",
    "    cands['model'] = label\n",
    "    candidates.append(cands)\n",
    "candidates = pd.concat(candidates)\n",
    "candidates.groupby('model').size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 680,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>candidate_hash</th>\n",
       "      <th>Alkaloid</th>\n",
       "      <th>NRP</th>\n",
       "      <th>Other</th>\n",
       "      <th>Polyketide</th>\n",
       "      <th>RiPP</th>\n",
       "      <th>Saccharide</th>\n",
       "      <th>Terpene</th>\n",
       "      <th>classes</th>\n",
       "      <th>Antibacterial</th>\n",
       "      <th>...</th>\n",
       "      <th>candidate_id</th>\n",
       "      <th>avg_prediction</th>\n",
       "      <th>nucl_length</th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>num_bio_domains</th>\n",
       "      <th>num_all_domains</th>\n",
       "      <th>num_proteins</th>\n",
       "      <th>protein_ids</th>\n",
       "      <th>pfam_ids</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>000749207dc6a8d25a7b67282a2fe375</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.940</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Saccharide</td>\n",
       "      <td>0.410000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP000970.1(1023496-1046922)</td>\n",
       "      <td>0.988859</td>\n",
       "      <td>23427</td>\n",
       "      <td>1023496</td>\n",
       "      <td>1046922</td>\n",
       "      <td>10</td>\n",
       "      <td>55</td>\n",
       "      <td>17</td>\n",
       "      <td>CP000970_941;CP000970_943;CP000970_944;CP00097...</td>\n",
       "      <td>PF13439;PF13692;PF00534;PF00535;PF13641;PF1011...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00244b39ef89b1f18ccf4da018657383</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0.250</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.609</td>\n",
       "      <td>0.01</td>\n",
       "      <td>Saccharide</td>\n",
       "      <td>0.583750</td>\n",
       "      <td>...</td>\n",
       "      <td>CP002727.1(1814667-1822576)</td>\n",
       "      <td>0.950768</td>\n",
       "      <td>7910</td>\n",
       "      <td>1814667</td>\n",
       "      <td>1822576</td>\n",
       "      <td>6</td>\n",
       "      <td>18</td>\n",
       "      <td>7</td>\n",
       "      <td>CP002727_1673;CP002727_1674;CP002727_1675;CP00...</td>\n",
       "      <td>PF03808;PF01050;PF00483;PF12804;PF07883;PF0231...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0024a1c619f0dba6e8420c60b6c815e7</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.92</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.030</td>\n",
       "      <td>0.030</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.710000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP010089.1(2566479-2595429)</td>\n",
       "      <td>0.970932</td>\n",
       "      <td>28951</td>\n",
       "      <td>2566479</td>\n",
       "      <td>2595429</td>\n",
       "      <td>4</td>\n",
       "      <td>28</td>\n",
       "      <td>7</td>\n",
       "      <td>CP010089_2573;CP010089_2574;CP010089_2575;CP01...</td>\n",
       "      <td>PF07690;PF05977;PF13347;PF00501;PF13193;PF0066...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>002d6748dad4145b1f09ca9dc93ead44</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.044</td>\n",
       "      <td>0.780</td>\n",
       "      <td>0.01</td>\n",
       "      <td>Saccharide</td>\n",
       "      <td>0.441667</td>\n",
       "      <td>...</td>\n",
       "      <td>CP002955.1(897167-922922)</td>\n",
       "      <td>0.979748</td>\n",
       "      <td>25756</td>\n",
       "      <td>897167</td>\n",
       "      <td>922922</td>\n",
       "      <td>8</td>\n",
       "      <td>33</td>\n",
       "      <td>16</td>\n",
       "      <td>CP002955_767;CP002955_768;CP002955_769;CP00295...</td>\n",
       "      <td>PF11721;PF13229;PF13524;PF13692;PF00534;PF0301...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>003041bd79286a337bdd94c332198c33</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.070</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.870</td>\n",
       "      <td>0.01</td>\n",
       "      <td>Saccharide</td>\n",
       "      <td>0.560000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP004372.1(1146582-1184447)</td>\n",
       "      <td>0.991482</td>\n",
       "      <td>37866</td>\n",
       "      <td>1146582</td>\n",
       "      <td>1184447</td>\n",
       "      <td>14</td>\n",
       "      <td>51</td>\n",
       "      <td>21</td>\n",
       "      <td>CP004372_1113;CP004372_1115;CP004372_1116;CP00...</td>\n",
       "      <td>PF02397;PF13579;PF13439;PF13692;PF00534;PF1347...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     candidate_hash  Alkaloid   NRP  Other  Polyketide   RiPP  \\\n",
       "0  000749207dc6a8d25a7b67282a2fe375      0.00  0.00   0.06       0.000  0.000   \n",
       "1  00244b39ef89b1f18ccf4da018657383      0.01  0.02   0.18       0.250  0.020   \n",
       "2  0024a1c619f0dba6e8420c60b6c815e7      0.00  0.92   0.02       0.030  0.030   \n",
       "3  002d6748dad4145b1f09ca9dc93ead44      0.00  0.04   0.05       0.076  0.044   \n",
       "4  003041bd79286a337bdd94c332198c33      0.01  0.00   0.03       0.070  0.010   \n",
       "\n",
       "   Saccharide  Terpene     classes  Antibacterial  \\\n",
       "0       0.940     0.00  Saccharide       0.410000   \n",
       "1       0.609     0.01  Saccharide       0.583750   \n",
       "2       0.000     0.00         NRP       0.710000   \n",
       "3       0.780     0.01  Saccharide       0.441667   \n",
       "4       0.870     0.01  Saccharide       0.560000   \n",
       "\n",
       "                         ...                          \\\n",
       "0                        ...                           \n",
       "1                        ...                           \n",
       "2                        ...                           \n",
       "3                        ...                           \n",
       "4                        ...                           \n",
       "\n",
       "                  candidate_id  avg_prediction  nucl_length nucl_start  \\\n",
       "0  CP000970.1(1023496-1046922)        0.988859        23427    1023496   \n",
       "1  CP002727.1(1814667-1822576)        0.950768         7910    1814667   \n",
       "2  CP010089.1(2566479-2595429)        0.970932        28951    2566479   \n",
       "3    CP002955.1(897167-922922)        0.979748        25756     897167   \n",
       "4  CP004372.1(1146582-1184447)        0.991482        37866    1146582   \n",
       "\n",
       "  nucl_end  num_bio_domains num_all_domains num_proteins  \\\n",
       "0  1046922               10              55           17   \n",
       "1  1822576                6              18            7   \n",
       "2  2595429                4              28            7   \n",
       "3   922922                8              33           16   \n",
       "4  1184447               14              51           21   \n",
       "\n",
       "                                         protein_ids  \\\n",
       "0  CP000970_941;CP000970_943;CP000970_944;CP00097...   \n",
       "1  CP002727_1673;CP002727_1674;CP002727_1675;CP00...   \n",
       "2  CP010089_2573;CP010089_2574;CP010089_2575;CP01...   \n",
       "3  CP002955_767;CP002955_768;CP002955_769;CP00295...   \n",
       "4  CP004372_1113;CP004372_1115;CP004372_1116;CP00...   \n",
       "\n",
       "                                            pfam_ids  \n",
       "0  PF13439;PF13692;PF00534;PF00535;PF13641;PF1011...  \n",
       "1  PF03808;PF01050;PF00483;PF12804;PF07883;PF0231...  \n",
       "2  PF07690;PF05977;PF13347;PF00501;PF13193;PF0066...  \n",
       "3  PF11721;PF13229;PF13524;PF13692;PF00534;PF0301...  \n",
       "4  PF02397;PF13579;PF13439;PF13692;PF00534;PF1347...  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 680,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deepbgc = pd.read_csv('../data/bacteria/candidates/{}/candidates_merged.csv'.format(DEEPBGC_PATH))\n",
    "deepbgc.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 924,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10926"
      ]
     },
     "execution_count": 924,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(deepbgc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 681,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Alkaloid</th>\n",
       "      <th>NRP</th>\n",
       "      <th>Other</th>\n",
       "      <th>Polyketide</th>\n",
       "      <th>RiPP</th>\n",
       "      <th>Saccharide</th>\n",
       "      <th>Terpene</th>\n",
       "      <th>Antibacterial</th>\n",
       "      <th>Cytotoxic</th>\n",
       "      <th>Inhibitor</th>\n",
       "      <th>Antifungal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Alkaloid  NRP  Other  Polyketide  RiPP  Saccharide  Terpene  Antibacterial  \\\n",
       "0         0    0      0           0     0           1        0              0   \n",
       "1         0    0      0           0     0           1        0              1   \n",
       "2         0    1      0           0     0           0        0              1   \n",
       "3         0    0      0           0     0           1        0              0   \n",
       "4         0    0      0           0     0           1        0              1   \n",
       "\n",
       "   Cytotoxic  Inhibitor  Antifungal  \n",
       "0          0          0           0  \n",
       "1          0          0           0  \n",
       "2          0          0           0  \n",
       "3          0          0           0  \n",
       "4          0          0           0  "
      ]
     },
     "execution_count": 681,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_classes = (deepbgc[['Alkaloid','NRP','Other','Polyketide','RiPP','Saccharide','Terpene','Antibacterial','Cytotoxic','Inhibitor','Antifungal']] >= 0.5).astype(np.int)\n",
    "all_classes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 682,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a39e77780>"
      ]
     },
     "execution_count": 682,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 1, figsize=(7, 4))\n",
    "sns.heatmap(all_classes.T.dot(all_classes), annot=True, fmt='.0f', cmap='Greys', ax=ax)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 683,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_regions(cands, highlight_cand=None, start=None, end=None):\n",
    "    if start:\n",
    "        cands = cands[cands['nucl_end'] >= start]\n",
    "    if end:\n",
    "        cands = cands[cands['nucl_start'] <= end]\n",
    "    fig = candidate_regions(cands, colors={m: c for m, c in zip(labels, colors)}, xlim=start or 0)\n",
    "    if highlight_cand is not None:\n",
    "        plt.axvspan(highlight_cand['nucl_start'], highlight_cand['nucl_end'], color='green', alpha=0.2)\n",
    "    return fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 684,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_candidate_region(candidate_hash, start=None, end=None):\n",
    "    cand = deepbgc[deepbgc['candidate_hash']==candidate_hash].iloc[0]\n",
    "    cands = candidates[candidates['contig_id']==cand['contig_id']]\n",
    "    return plot_regions(cands, highlight_cand=cand, start=start, end=end)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DeepBGC Novel candidates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 685,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BGC_ID</th>\n",
       "      <th>num_proteins</th>\n",
       "      <th>classes</th>\n",
       "      <th>source</th>\n",
       "      <th>species</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contig_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BGC0000001.1</th>\n",
       "      <td>BGC0000001</td>\n",
       "      <td>28</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>Verrucosispora maris AB-18-032</td>\n",
       "      <td>Verrucosispora</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000002.1</th>\n",
       "      <td>BGC0000002</td>\n",
       "      <td>34</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>Kutzneria albida DSM 43870</td>\n",
       "      <td>Kutzneria</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000003.1</th>\n",
       "      <td>BGC0000003</td>\n",
       "      <td>6</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>Alternaria alternata</td>\n",
       "      <td>Alternaria</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000004.1</th>\n",
       "      <td>BGC0000004</td>\n",
       "      <td>23</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>Aspergillus oryzae</td>\n",
       "      <td>Aspergillus</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000005.1</th>\n",
       "      <td>BGC0000005</td>\n",
       "      <td>1</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>Aspergillus parasiticus</td>\n",
       "      <td>Aspergillus</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  BGC_ID  num_proteins     classes  \\\n",
       "contig_id                                            \n",
       "BGC0000001.1  BGC0000001            28  Polyketide   \n",
       "BGC0000002.1  BGC0000002            34  Polyketide   \n",
       "BGC0000003.1  BGC0000003             6  Polyketide   \n",
       "BGC0000004.1  BGC0000004            23  Polyketide   \n",
       "BGC0000005.1  BGC0000005             1  Polyketide   \n",
       "\n",
       "                                      source         species  \n",
       "contig_id                                                     \n",
       "BGC0000001.1  Verrucosispora maris AB-18-032  Verrucosispora  \n",
       "BGC0000002.1      Kutzneria albida DSM 43870       Kutzneria  \n",
       "BGC0000003.1            Alternaria alternata      Alternaria  \n",
       "BGC0000004.1              Aspergillus oryzae     Aspergillus  \n",
       "BGC0000005.1         Aspergillus parasiticus     Aspergillus  "
      ]
     },
     "execution_count": 685,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mibig = pd.read_csv('../data/mibig/mibig_properties.csv').set_index('contig_id')\n",
    "mibig.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 797,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "566\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>candidate_hash</th>\n",
       "      <th>Alkaloid</th>\n",
       "      <th>NRP</th>\n",
       "      <th>Other</th>\n",
       "      <th>Polyketide</th>\n",
       "      <th>RiPP</th>\n",
       "      <th>Saccharide</th>\n",
       "      <th>Terpene</th>\n",
       "      <th>classes</th>\n",
       "      <th>Antibacterial</th>\n",
       "      <th>...</th>\n",
       "      <th>candidate_id</th>\n",
       "      <th>avg_prediction</th>\n",
       "      <th>nucl_length</th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>num_bio_domains</th>\n",
       "      <th>num_all_domains</th>\n",
       "      <th>num_proteins</th>\n",
       "      <th>protein_ids</th>\n",
       "      <th>pfam_ids</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0289639a11eb66fdf05558583f9a45a3</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>0.194961</td>\n",
       "      <td>0.259762</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.259331</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP004886.1(1996131-2060701)</td>\n",
       "      <td>0.952772</td>\n",
       "      <td>64571</td>\n",
       "      <td>1996131</td>\n",
       "      <td>2060701</td>\n",
       "      <td>1</td>\n",
       "      <td>77</td>\n",
       "      <td>53</td>\n",
       "      <td>CP004886_2410;CP004886_2411;CP004886_2412;CP00...</td>\n",
       "      <td>PF03007;PF03007;PF06974;PF03007;PF16817;PF0190...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5a2182c16f4ab87f299bf25f31225f87</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.208003</td>\n",
       "      <td>0.305182</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.241884</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.390000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003494.1(1693261-1753880)</td>\n",
       "      <td>0.961767</td>\n",
       "      <td>60620</td>\n",
       "      <td>1693261</td>\n",
       "      <td>1753880</td>\n",
       "      <td>1</td>\n",
       "      <td>70</td>\n",
       "      <td>48</td>\n",
       "      <td>CP003494_1606;CP003494_1607;CP003494_1608;CP00...</td>\n",
       "      <td>PF00934;PF03007;PF06974;PF16817;PF01906;PF0184...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>f3f2b5cb55121b1bea879afa884e3ab3</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Saccharide</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP008802.1(946248-984397)</td>\n",
       "      <td>0.985270</td>\n",
       "      <td>38150</td>\n",
       "      <td>946248</td>\n",
       "      <td>984397</td>\n",
       "      <td>15</td>\n",
       "      <td>90</td>\n",
       "      <td>25</td>\n",
       "      <td>CP008802_802;CP008802_803;CP008802_804;CP00880...</td>\n",
       "      <td>PF12804;PF00483;PF04321;PF01370;PF13460;PF0107...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>f1ca8035648e3be253f1e5ee8e52459c</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.074303</td>\n",
       "      <td>0.350000</td>\n",
       "      <td>0.095697</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.403500</td>\n",
       "      <td>...</td>\n",
       "      <td>CP006936.1(3460480-3493402)</td>\n",
       "      <td>0.963722</td>\n",
       "      <td>32923</td>\n",
       "      <td>3460480</td>\n",
       "      <td>3493402</td>\n",
       "      <td>2</td>\n",
       "      <td>32</td>\n",
       "      <td>25</td>\n",
       "      <td>CP006936_3288;CP006936_3289;CP006936_3291;CP00...</td>\n",
       "      <td>PF02405;PF12680;PF07366;PF16859;PF00440;PF1397...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>53d05d64e99c1b41d78210f57b6201e6</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.245500</td>\n",
       "      <td>0.129833</td>\n",
       "      <td>0.544667</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003233.1(2146676-2176897)</td>\n",
       "      <td>0.957012</td>\n",
       "      <td>30222</td>\n",
       "      <td>2146676</td>\n",
       "      <td>2176897</td>\n",
       "      <td>8</td>\n",
       "      <td>35</td>\n",
       "      <td>23</td>\n",
       "      <td>CP003233_2275;CP003233_2278;CP003233_2279;CP00...</td>\n",
       "      <td>PF01266;PF00890;PF14081;PF00141;PF01475;PF0197...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     candidate_hash  Alkaloid       NRP     Other  Polyketide  \\\n",
       "0  0289639a11eb66fdf05558583f9a45a3      0.04  0.040000  0.194961    0.259762   \n",
       "1  5a2182c16f4ab87f299bf25f31225f87      0.01  0.010000  0.208003    0.305182   \n",
       "2  f3f2b5cb55121b1bea879afa884e3ab3      0.00  0.010000  0.040000    0.030000   \n",
       "3  f1ca8035648e3be253f1e5ee8e52459c      0.00  0.074303  0.350000    0.095697   \n",
       "4  53d05d64e99c1b41d78210f57b6201e6      0.02  0.245500  0.129833    0.544667   \n",
       "\n",
       "   RiPP  Saccharide   Terpene     classes  Antibacterial  \\\n",
       "0  0.30        0.04  0.259331         NaN       0.360000   \n",
       "1  0.33        0.01  0.241884         NaN       0.390000   \n",
       "2  0.01        0.91  0.000000  Saccharide       0.600000   \n",
       "3  0.29        0.04  0.170000         NaN       0.403500   \n",
       "4  0.06        0.06  0.040000  Polyketide       0.333333   \n",
       "\n",
       "                         ...                          \\\n",
       "0                        ...                           \n",
       "1                        ...                           \n",
       "2                        ...                           \n",
       "3                        ...                           \n",
       "4                        ...                           \n",
       "\n",
       "                  candidate_id  avg_prediction  nucl_length nucl_start  \\\n",
       "0  CP004886.1(1996131-2060701)        0.952772        64571    1996131   \n",
       "1  CP003494.1(1693261-1753880)        0.961767        60620    1693261   \n",
       "2    CP008802.1(946248-984397)        0.985270        38150     946248   \n",
       "3  CP006936.1(3460480-3493402)        0.963722        32923    3460480   \n",
       "4  CP003233.1(2146676-2176897)        0.957012        30222    2146676   \n",
       "\n",
       "  nucl_end  num_bio_domains num_all_domains num_proteins  \\\n",
       "0  2060701                1              77           53   \n",
       "1  1753880                1              70           48   \n",
       "2   984397               15              90           25   \n",
       "3  3493402                2              32           25   \n",
       "4  2176897                8              35           23   \n",
       "\n",
       "                                         protein_ids  \\\n",
       "0  CP004886_2410;CP004886_2411;CP004886_2412;CP00...   \n",
       "1  CP003494_1606;CP003494_1607;CP003494_1608;CP00...   \n",
       "2  CP008802_802;CP008802_803;CP008802_804;CP00880...   \n",
       "3  CP006936_3288;CP006936_3289;CP006936_3291;CP00...   \n",
       "4  CP003233_2275;CP003233_2278;CP003233_2279;CP00...   \n",
       "\n",
       "                                            pfam_ids  \n",
       "0  PF03007;PF03007;PF06974;PF03007;PF16817;PF0190...  \n",
       "1  PF00934;PF03007;PF06974;PF16817;PF01906;PF0184...  \n",
       "2  PF12804;PF00483;PF04321;PF01370;PF13460;PF0107...  \n",
       "3  PF02405;PF12680;PF07366;PF16859;PF00440;PF1397...  \n",
       "4  PF01266;PF00890;PF14081;PF00141;PF01475;PF0197...  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 797,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novel = pd.read_csv('../data/bacteria/candidates/{}/novel.csv'.format(DEEPBGC_PATH))\n",
    "print(len(novel))\n",
    "novel.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 798,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "227"
      ]
     },
     "execution_count": 798,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 687,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = novel['nucl_length'].plot.box(showfliers=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 688,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = novel[['num_all_domains','num_bio_domains']].plot.box(showfliers=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 689,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Alkaloid</th>\n",
       "      <th>NRP</th>\n",
       "      <th>Other</th>\n",
       "      <th>Polyketide</th>\n",
       "      <th>RiPP</th>\n",
       "      <th>Saccharide</th>\n",
       "      <th>Terpene</th>\n",
       "      <th>Antibacterial</th>\n",
       "      <th>Cytotoxic</th>\n",
       "      <th>Inhibitor</th>\n",
       "      <th>Antifungal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Alkaloid  NRP  Other  Polyketide  RiPP  Saccharide  Terpene  Antibacterial  \\\n",
       "0         0    0      0           0     0           0        0              0   \n",
       "1         0    0      0           0     0           0        0              0   \n",
       "2         0    0      0           0     0           1        0              1   \n",
       "3         0    0      0           0     0           0        0              0   \n",
       "4         0    0      0           1     0           0        0              0   \n",
       "\n",
       "   Cytotoxic  Inhibitor  Antifungal  \n",
       "0          0          0           0  \n",
       "1          0          0           0  \n",
       "2          0          0           0  \n",
       "3          0          0           0  \n",
       "4          0          0           0  "
      ]
     },
     "execution_count": 689,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novel_all_classes = (novel[['Alkaloid','NRP','Other','Polyketide','RiPP','Saccharide','Terpene','Antibacterial','Cytotoxic','Inhibitor','Antifungal']] >= 0.5).astype(np.int)\n",
    "novel_all_classes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 690,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a3b0c8a90>"
      ]
     },
     "execution_count": 690,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 1, figsize=(7, 4))\n",
    "sns.heatmap(novel_all_classes.T.dot(novel_all_classes), annot=True, fmt='.0f', cmap='Greys', ax=ax)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Biosynthetic class distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 917,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_class_counts(classes, colors, group_hybrids=False, title='', ax=None):\n",
    "    classes = classes.fillna('No confident class')\n",
    "    if group_hybrids:\n",
    "        classes = classes.apply(lambda c: 'Hybrids' if ';' in c else c)\n",
    "    class_counts = classes.value_counts()\n",
    "    c = [colors.get(c, 'white') for c in class_counts.index]\n",
    "    total = class_counts.sum()\n",
    "    l = ['{} ({:.0%})'.format(c,v/total) for c,v  in class_counts.items()]\n",
    "    ax = class_counts.plot.pie(wedgeprops=dict(width=0.5), ax=ax, colors=c, labels=l)\n",
    "    ax.set_ylabel('')\n",
    "    ax.set_xlabel(title)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 856,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_species_counts(species, title='', ax=None):\n",
    "    species = species.apply(lambda species: species.split('_')[0])\n",
    "    top_species = species.value_counts()[:10].index\n",
    "    top_species_counts = species.apply(lambda s: s if s in top_species else 'Others').value_counts()\n",
    "    c = [SPECIES_COLORS.get(c, 'lightgrey') for c in top_species_counts.index]\n",
    "    total = top_species_counts.sum()\n",
    "    l = ['{} ({:.0%})'.format(c,v/total) for c,v  in top_species_counts.items()]\n",
    "    ax = top_species_counts.plot.pie(wedgeprops=dict(width=0.5), colors=c, labels=l, ax=ax)\n",
    "    ax.set_ylabel('')\n",
    "    ax.set_xlabel(title)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Activity class distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 694,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_class_counts(deepbgc['classes1'], ACT_COLORS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 693,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_class_counts(novel['classes1'], ACT_COLORS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 923,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x576 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(2, 2, figsize=(13, 8))\n",
    "plot_class_counts(deepbgc['classes'], BIO_COLORS, group_hybrids=True, ax=axes[0][0])\n",
    "plot_class_counts(novel['classes'], BIO_COLORS, group_hybrids=True, ax=axes[0][1])\n",
    "\n",
    "plot_species_counts(deepbgc['species'], title='All candidates', ax=axes[1][0])\n",
    "plot_species_counts(novel['species'], title='Novel candidates', ax=axes[1][1])\n",
    "\n",
    "fig.savefig('../data/figures/paper/supplementary/10a.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 925,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_heatmap(heatmap_cands, columns=None, **kwargs):\n",
    "    heatmap = pd.DataFrame()\n",
    "    top_species = species.value_counts()[:10].index\n",
    "    heatmap['species'] = heatmap_cands['species'].apply(lambda species: species.split('_')[0]).apply(lambda s: s if s in top_species else 'Others')\n",
    "    heatmap['classes'] = heatmap_cands['classes'].fillna('No confident class')\n",
    "    heatmap = heatmap.pivot_table(values='species', index='species', columns=['classes'], aggfunc=len)\n",
    "    if columns is not None:\n",
    "        heatmap = heatmap.transpose().loc[columns].fillna(0).transpose()\n",
    "    heatmap = heatmap.fillna(0).astype(np.int)\n",
    "    heatmap = heatmap.loc[top_species.tolist() + ['Others']]\n",
    "    heatmap = heatmap.apply(lambda row: row / row.sum(), axis=1)\n",
    "    sns.heatmap(heatmap, annot=True, fmt='.0%', cmap='Greys', **kwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 927,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 936x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 2, figsize=(13, 5))\n",
    "columns = deepbgc['classes'].sort_values().fillna('No confident class').unique()\n",
    "plot_heatmap(deepbgc, columns=columns, ax=axes[0], cbar=False)\n",
    "plot_heatmap(novel, columns=columns, ax=axes[1], cbar=False)\n",
    "axes[1].set_ylabel('')\n",
    "plt.tight_layout()\n",
    "fig.savefig('../data/figures/paper/supplementary/10b.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 935,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/prihodad/anaconda3/lib/python3.6/site-packages/numpy/core/fromnumeric.py:51: FutureWarning: reshape is deprecated and will raise in a subsequent release. Please use .values.reshape(...) instead\n",
      "  return getattr(obj, method)(*args, **kwds)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0.98, '')"
      ]
     },
     "execution_count": 935,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEICAYAAACavRnhAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAGVlJREFUeJzt3XuUZWV95vHvwzU013GBHRuRziAgiKihQYJLbZJIkEsEL+gEMMg4LUFXZBwvmImXqAsxJiM6C5H20kiU0QEVFBokEQovGBSBRuQygLbAmChG0to9iDb85o+9Cw5ldddp3iqqquv7WeusdfZ+9373u88+dZ79vvvUPqkqJElqscl0N0CSNPsZJpKkZoaJJKmZYSJJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmm023Q1oseOOO9bChQunuxlTbs2aNWy99dbT3QxNAo/lxmOuHMvvfve7P6uqnSZablaHycKFC7n22munuxlTbmRkhMWLF093MzQJPJYbj7lyLJP8aJjlHOaSJDUzTCRJzQwTSVIzw0SS1MwwkSQ1M0wkSc0ME0lSM8NEktRsVv/ToiRNlSSTUk9VTUo9M509E0kaR1Wt97HrWy+ecJm5EiRgmEiSJoFhIklqZphIkpoZJpKkZoaJJKmZYSJJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmhkmkqRmhokkqZlhIklqZphIkpoZJpKkZoaJJKmZYSJJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmg0VJkmOTnJjkhuSXJFktySbJjkjya1J7khy0sDyuyf5WpKbk3w7ydMGyk7s59+e5Kwkm/fz5yU5L8ktSW5LctTk764kaSpMGCZJtgI+Dbykqp4FfBn4MPBaYA9gH2B/4JQkB/SrfQb4aFXtDbwTuCCdfYC/AV4A7AnsAPzXfp13Aaurai/ghcCZSZ48KXspSZpSw/RMNgUCbN9PbwP8CjgaWFZVa6vqPuCzwHFJdgae1k9TVZf26zwbeDHwpaq6t6oeAs4GjuvrPRr4WL/OXcA/Asc076EkacptNtECVbW6H8K6Osm/0YXLc4GLgbsHFr0H2BfYBfhxHxaDZU/uy1aOM5++bGx9v9UzSbIEWAIwf/58RkZGJtqFWW/16tVzYj/nAo/lxsVj+YgJwyTJM4B3AHtX1Z1J/hL4PF2o1OCiwIN0vZ0aW806ykbnM0HZw6pqKbAUYNGiRbV48eKJdmHWGxkZYS7s51zgsdyIXHaJx3LAMMNcfwJ8s6ru7KfPpLtO8iNgwcByC+h6E3cBT0qSdZSNtw4TlEmSZrBhwuQ64AVJ5vfTRwE/BC4CTkyyWZIdgFcCF1bVPcAdwCsAkvwJ8BDwPeBLwJ8meWIfNkuAC/t6L+qn6S+8H0o3lCZJmuGGuWZyRZIPACNJfg38nO5C+m3AbsAKYAvg7Kq6ql/tPwEfS/LXdBfrX95fQ7kxybuBK4DNgWuA9/frvBM4K8n36YbQ3jzQG5IkzWAThglAVZ1JN7w11inrWP52YPE6ypYBy8aZvxo4fpj2SJJmFv8DXpLUzDCRJDUzTCRJzQwTSVIzw0SS1MwwkSQ1M0wkSc0ME0lSM8NEktTMMJEkNTNMJEnNDBNJUjPDRJLUzDCRJDUzTCRJzQwTSVIzw0SS1MwwkSQ1M0wkSc0ME0lSM8NEktTMMJEkNTNMJEnNDBNJUjPDRJLUzDCRJDUzTCRJzQwTSVIzw0SS1MwwkSQ1M0wkSc0ME0lSM8NEktRsqDBJ8owkI0muT3Jtkv36+W9LcmuSO5K8K0n6+TsluTTJzUluSnLQQF2HJ7kxyW1Jzk+yXT9/0yRnDNR30lTssCRp8k0YJknmAZcDf1tVzwbeA3wmyWHAMcB+wD7AwcDL+9XOBL5eVXsDxwHnJ5mXZCdgGfDSqtoT+AFwer/Oa4E9+rr2B05JcsDk7KYkaSoN0zM5BLizqpb301+iC5GjgfOqak1V/YouJI5LshlwBPAxgKq6AbgdOLSv6ztVdXtf11nAsX2P5mhgWVWtrar7gM/SBZEkaYbbbIhl9gD+NckngGcC/w68BdgF+OrAcvcATwZ2BDapqnvHKZsH3D1m/nbAtn19Y8v2HduYJEuAJQDz589nZGRkiF2Y3VavXj0n9nMu8FhuXDyWjxgmTDYHDgMOrqprkrwYWA7cAtTAcgEepOvt1Jg61lfGOspG13mUqloKLAVYtGhRLV68eIhdmN1GRkaYC/u5MegvGzarGu/PRDPKZZf4dzlgmGGuHwO3VNU1AFV1EbAp8BCwYGC5BXS9iZ8CSfKEccruGrPOzsB9VbVmnLLRdaRZo6rW+9j1rRdPuIxBotlomDC5FPi9gW9wPZ+uB3EG3fWOrZNsCZwAXFhVa4FL6IeikuwL7A2M0F3IPzDJ7n3dJwEX9c8vAk5MslmSHYBXAhc276EkacpNOMxVVf+a5CjgI0m2Bh4AXlJV30jyDODbwBZ0YXBuv9rJwMeT3EQXPMdX1SqAJK8GLkiyBXAn8Kp+nbOA3YAVfX1nV9VVk7SfkqQpNMw1E6rqa8Bzxpl/GnDaOPN/Ahy5jrqW011zGTt/LXDKMO2RJM0s/ge8JKmZYSJJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmhkmkqRmhokkqZlhIklqZphIkpoZJpKkZoaJJKmZYSJJamaYSJKaDfV7JpK0MXnm31zOqvt/01zPwlMvaa5j+602Z8U7D2muZ7oZJpLmnFX3/4aVpx/eVMfIyAiLFy9ubstkBNJM4DCXJKmZYSJJauYw1zRLMml1VdWk1SVJG8KeyTSrqgkfu7714qGWk6TpYphIkpoZJpKkZoaJJKmZYSJJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmhkmkqRmhokkqZlhIklqNnSYJDkqyS8Hpt+W5NYkdyR5V/rb3ybZKcmlSW5OclOSgwbWOTzJjUluS3J+ku36+ZsmOWOgvpMmcyclSVNrqDBJsjvwd8BoYBwGHAPsB+wDHAy8vF/8TODrVbU3cBxwfpJ5SXYClgEvrao9gR8Ap/frvBbYo69rf+CUJAe0754k6fEwYZgkmQd8GnjjwOyjgfOqak1V/YouJI5LshlwBPAxgKq6AbgdOBQ4BPhOVd3e13EWcGzfozkaWFZVa6vqPuCzdEEkSZoFhvlxrLP7x40D83YBvjowfQ/wZGBHYJOqunecsnnA3WPmbwds29c3tmzf8RqTZAmwBGD+/PmMjIwMsQuz31zZz7nAYzkztB6H1atXT9qx3BjeE+sNkyQnA2ur6pNJFg4UbQIM/hpTgAfHmT9RGesoG13nt1TVUmApwKJFi2rx4sXr24WNw2WXMCf2cy7wWM4Mk3AcRkZGJudYbiTviYl6JicA85LcAGwBbNU/vw5YMLDcArrexE+BJHlCVf18TNkq4DkD6+wM3FdVa5LctY76JEmzwHqvmVTVAVW1T1U9CzgMuL9//kW66x1bJ9mSLnQurKq1wCX0w1BJ9gX2BkaAy4ED+4v5ACcBF/XPLwJOTLJZkh2AVwIXTt5uSpKm0jDXTH5LVX05yTOAb9P1WC4Czu2LTwY+nuQmuqGr46tqFUCSVwMXJNkCuBN4Vb/OWcBuwIq+vrOr6qrHtkuSpMfb0GFSVSuBbQamTwNOG2e5nwBHrqOO5cDyceavBU4Zti2SpJnF/4CXJDUzTCRJzQwTSVIzw0SS1MwwkSQ1M0wkSc0ME0lSs8f0T4uSNJttu9epPONTp7ZX9KnJaAvA4e0VTTPDRNKc88tbTmfl6W0f4JN1o8eFp17SXMdM4DCXJKmZYSJJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmhkmkqRmhokkqZlhIklqZphIkpoZJpKkZt7oURrSM//mclbd/5vmeibjxn7bb7U5K955SHM90mQxTKQhrbr/N95pVloHh7kkSc0ME0lSM8NEktTMMJEkNfMC/BSbKd8A8ts/kqaSYTLFZso3gPz2j6Sp5DCXJKmZYSJJamaYSJKaGSaSpGZDhUmS45KsSHJDkquTLOrnvy3JrUnuSPKuJOnn75Tk0iQ3J7kpyUEDdR2e5MYktyU5P8l2/fxNk5wxUN9JU7HDkqTJN2GYJNkT+ABwaFU9C3gv8IUkhwHHAPsB+wAHAy/vVzsT+HpV7Q0cB5yfZF6SnYBlwEurak/gB8Dp/TqvBfbo69ofOCXJAZOzm5KkqTTMV4MfAF5TVf/ST18L/C5dcJxXVWsAkiwDjkvyBeAI4HUAVXVDktuBQ4GtgO9U1e19XWcBK5K8DjgaWFpVa4H7knyWLoi+PQn7KUmPMilfl79scu4AvTGYMEyqaiWwEqAfxvofwJeAJwFfGVj0HuDJwI7AJlV17zhl84C7x8zfDtgW2GWcsn3HtifJEmAJwPz58xkZGZloF6ZdaxtXr149Kfs5G16rmW6mHMvJaMtcds6hWzfXccJlayalHtg4juXQ/7SYZGvgHLoP/UOB/w3U4CLAg3RDZzV29fWUsY6y0XUepaqWAksBFi1aVJNxO+8pddklzf9wOCm3LZ+Edsx5M+VYTlJb1Mhj8CjDXoB/CnA13Yf7wVX178BdwIKBxRbQ9SZ+2q2SJ4xTNnadnYH7+qGyddUnSZrhhrkAvy0wAnyhql5ZVff3RRcBxybZOsmWwAnAhf01j0voh6KS7Avs3ddxOXBgkt37Ok7q6xmt78QkmyXZAXglcGHzHkqSptwww1yvB3YFjk5y9MD8PwK+QHeBfAu6MDi3LzsZ+HiSm+iGro6vqlUASV4NXJBkC+BO4FX9OmcBuwEr+vrOrqqrGvZNkvQ4GeYC/PuA962j+LT+MXadnwBHrqO+5cDyceavBU6ZqD2SpJnH/4CXJDUzTCRJzQwTSVIzfxxrim2716k841Ontlf0qdZ2ALT9SJckrYthMsV+ecvp/tKipI2ew1ySpGaGiSSpmcNc0pBmyvWvri3gNTDNJIaJNKSZcv0LvAammcdhLklSM8NEktTMMJEkNTNMJEnNDBNJUjPDRJLUzDCRJDUzTCRJzfynxcfBpPyD2WVtdWy/1ebtbZCkdTBMpljrf0xDF0aTUY8kTRWHuSRJzQwTSVIzw0SS1MwwkSQ1M0wkSc38Npe0AWbC17zBr3o/HpJMvMz7J66nqiahNTOfYSINya95zy0ThcBk/dDZxsJhLklSM8NEktTMMJEkNTNMJEnNDBNJUjPDRJLUzK8GT7NhvssOfp9d0sw2o3omSQ5PcmOS25Kcn2S76W7TVKuqCR9XXnnlUMtJ0nSZMWGSZCdgGfDSqtoT+AFw+vS2SpI0jBkTJsAhwHeq6vZ++izg2Aw7DiRJmjYz6ZrJLsDdA9P3ANsB2wK/GJ2ZZAmwBGD+/PmMjIw8jk2cHqtXr54T+7kxOPjggydcZpjrX1deeeUktEZTyb/LR5tJYbIJMN7A/4ODE1W1FFgKsGjRopoL98bxHkCzh/dzmjs8lo82k4a57gIWDEzvDNxXVWumqT2SpCHNpDC5HDgwye799EnARdPYHknSkGbMMFdV/TTJq4ELkmwB3Am8apqbJUkawowJE4CqWg4sn+52SJI2zEwa5pIkzVKGiSSpmWEiSWpmmEiSmmU23yAwyb3Aj6a7HY+DHYGfTXcjNCk8lhuPuXIsd62qnSZaaFaHyVyR5NqqWjTd7VA7j+XGw2P5aA5zSZKaGSaSpGaGyeywdLoboEnjsdx4eCwHeM1EktTMnokkqZlhIklqZpisR5JNk7wxybVJbkhyc5L3J9kyyTlJ3tRQ9+VJdnwM661MclvfntHH8r5seZK9N7C+i5OcsKHt2NgkWZjkwYHX9MYkVyc5Zoq2d0KSVf22VvTvrYuTPGlgmXlJ3pvk+/3jh0n+YXCZfrk/T/Ktgffo0iQ7TEW7NxZJ3pHkxf3zdyd5Vf/8SUk+l+R7/XvgmtHl+vKVSdYk2WZMfSckqSQvGzP/80l+lmTemPkTbWdknLoWJlk9ea/C5JpRdw2egc4C/gPwR1W1KsnWwGeAjzPmFyAfgxc2rHtsVV07dmZVHdZQp+D+qnrW6ESSXYGvJnmwqj4/Bdv7elUdMbC9jwDvBv5Lkk2By4DvAwdU1ZokmwBvBi5N8uyqqiR/BbwIOKqqfpJkc+AM4MvA86agzRuLPwRuBqiqdwzM/zjwT1X1CoD+5OybSQ6qqlv6ZX4GvAQ4d2C9VwE/GdxAkgXA84F/7ss/uoHbmVXsmaxDkoXAscB/rqpVAP2vPp4EXDhm2RrsZYxOJ9kmyfn9GeN1ST6WZJMky/pFr0yyS5Kdk3wxyXf7s5S/Gm1Dkrv7Xsz/GXtGOk6bVyZZlGRxkm/2Z7HXJ7kpyXP7ZRYk+cf+THc58LsD6+/Vb+u7fZtP7Ocv7s+er+7bt2Xbqzs7VNWPgHcAb06yRZIP9sdxRbqe6XYAExy/HyU5e6AHMu4HfB8C2wH/0s86CtgBeN3or41W1UNV9X7gs8A2/cnN24ATq+on/TK/oQucj6b7XaCNXv839aH+7P7mJLckeW5/jD6c5MokdyT5Qv83+TpgEfCBJEfn0aMMTwK26oObqroZ+FPgvoFNfho4bmD7uwLbALeOadoS4KvAOcAbkmSgbJjtzC5V5WOcB/BS4NvrKT8HeFP/vIAdB8qK7lYLxwOX9fM2BT4GPHXsOsAVwJH989/pp48BFvbLPW+g7pXAbcANA49nDZQtAhYDawfm/zfgqv75F4H39M+fCqwGTqDrpX4f+P2+bHu6M7cD+/oepLutwrQfmyk63guB1ePMfzqwhi5UPsAj34A8DfjIkMfvz/qyFwE/BjbvX/NV/fFbQXe2e8/oa0zXu/j7Cdq8H/DT6X7tpvsB/AFwPrBJP30qXc/sHOCbwJb9a34d8Op+mRHgZf3zwb/lP+yP0c/ofun1zcDOA9taCRxE1wt5Uj/vr4HXj6lzM+D/Akf02/858KKBeibazgjwQx79d37zeO/RmfKwZ7JuD9Hec/sG8PQkI3Rv8DOq6o7BBfqzyxcA70lyA12X+CnA6HDLWuBbY+o9tqqeNfC4YZxt/2hg/nXAE/rnf0z3x0Pfliv6+XsAuwGf7NtxFbAV8Oy+/O7qztTnmgL+H92HwouB6/vX5yhg7yGO331VdR5AVV1KF8r79mVf74/fM4EnAmcDl/VnsOm3DUCSg/PI9Zy7khzB5LxHZ72q+hbdB/prk/wd8DK6ngJ0J3MPVNdj+x6P/B2sq64r6I7fUcA1wJHArUn2H1js18AFwJ/1068AzhtT1YvpTiAvq6oH6HqTp2zgdt48+HcOzOhh7Dn/RlyPa4C9kmw7OLMf0riE7oP2UUV9+cNDC1X1Q7qz//fRDWH8U5Ijx6y3ab/uQQNvmgPpznwBHqiqtY+h/fcPPK/R9o15Dl1YjbZj1Zg374HA6JDcjL3wN8X2p/sQ2hR4w8BrcwDdh9ZEx2/ssduEca63VdVDwP8EnkYXLFfThdRo+ZUD9f+Argd0M7B5kt0H60ryO+m+jLGgbddnhySHA5f0kxfRXZsYfY+v6+9gvHqe2F+3qqr6RlWdVlXPBz4H/PmYxc8FjktyEHBbVf18TPnJdJ8RdyRZSRcahyR5+gZuZ9YwTNahqn5Md7H9kwNj49sBHwH+jUe/Se+lG16CR85WSPIXdB/Gl1fVW4GvAL/fFz8IbF5Vv6A7m31jv84OdF3zh7/ZMckuoxvLJclTgIP7+bcB9yc5ri/bBbiJbihlTkqyB/B24O/pjt3r+2snm9ANWb5viOO3U5JD+7IjgdEz5PEcTjeMci/weWBNkjMy8M2hJM8B/iPwYH/G+37gE0nm9+VbAh8Etu7fw3PBC4EvV9VZwLV0H9ybTrDOWrqhr0E/7+t6+PpGum9h7UbXu39YVV1DFxan0ff0R/XvmxcA+1XVwv6xAPga8IYN2c5sYpis38l0Z39X90MY1/TTrxmz3F8CZya5DtiLRy6inkv3pr45yXfprkN8uC87H7gqyT50AXRgku/12/hfVfWZKdqn19ENz9wCfIJuLJaq+jXdB+BrktwIXA68vaq+OUXtmIm2GhhKuo7uQ+JtVXUJ8B66D/rr6d4DobsWBes/fr8Cjk+yAvjvdN+6Gu2ZPK/f1vVJbqIL+aOru9C+FjiU7sTla6PDW8B76cbXLwSoqtPogucr/Xt0Rd+2qToZmYk+CizuX//rgDuB32P9n29fAt6X5OGeQP+aH0J3DeaH/TG5Bri4qj45Th3/AOxJd4I26C+AL44d0qb7pt7xdF+s2JDtzAreTkWaIum+EXhTVW0zwaLSrGfPRJLUzJ6JJKmZPRNJUjPDRJLUzDCRJDUzTKRpliHu3Jzu/mg3PU5NkjaYYSJJamaYSBuo7yV8K93vUdyQ7g7NR6a7G/NdST7YL7ck3R2bV6S7G/Me/fwNvnOzNNP5eybSY7M/cHJVXZ/kUrpbwS+muwfbj5N8B3gL8AdVdW8/jHVhkqcDZwL/XFVvT/JU+rsQJNmM7gaCx1fVdUm2B76V5ObHe+ekDWWYSI/ND6vq+v75nXQ3yfw18LMkv6C7CeTnqupegKo6J8mH6G5L/8fAm/r5dyQZ787No9sZvXPzrPzBJM0dhon02DwwZvo3Y6aLgVvI90J3c8EJ79z88ArdDRxX0d2JWJqxvGYiTY0R4JVJdgJI8mq6mzbegXdu1kbInok0Na6k+/GqK/pb1t8LHFFVD6X72dhl/Z2b72Hgzs1JXgx8KMlb6Hoxb6+qbyZZPC17IQ3Je3NJkpo5zCVJamaYSJKaGSaSpGaGiSSpmWEiSWpmmEiSmhkmkqRm/x/ufKi3iz6IRgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = candidates.boxplot('nucl_length', by='model', showfliers=False)\n",
    "ax.set_title('')\n",
    "ax.figure.suptitle('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 966,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_coverage_distribution(coverage, figsize=(10, 2), cov_lim=None, cov_step=100, ax=None):\n",
    "    dist = []\n",
    "    thresholds = np.arange(0, 1, 0.001)\n",
    "    for threshold in thresholds:\n",
    "        num_over_threshold = coverage[coverage['coverage'] > threshold].groupby('model').size()\n",
    "        dist.append(num_over_threshold)\n",
    "    dist = pd.DataFrame(dist, index=thresholds)\n",
    "\n",
    "    for i, name in enumerate(dist.columns):\n",
    "        color = MODEL_COLORS.get(name, 'grey')\n",
    "        is_postproc = 'postproc' in name\n",
    "        linestyle = '--' if is_postproc else '-'\n",
    "        ax.plot(dist[name], color=color, linestyle=linestyle, lw=1)\n",
    "        \n",
    "    if cov_lim:\n",
    "        ax.set_yticks(list(range(0, cov_lim, cov_step)))\n",
    "        ax.set_ylim([0, cov_lim])\n",
    "    ax.set_xlim([-0.01,1.01])\n",
    "    ax.set_xticklabels(['>{:.0f}%'.format(x*100) for x in ax.get_xticks()]);\n",
    "    ax.set_ylabel('# BGCs')\n",
    "    ax.set_xlabel('BGC Coverage')\n",
    "    plt.tight_layout()\n",
    "    return fig"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 989,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coverage</th>\n",
       "      <th>any_exact</th>\n",
       "      <th>max_covered</th>\n",
       "      <th>contig_id</th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>nucl_length</th>\n",
       "      <th>candidate_id</th>\n",
       "      <th>classes</th>\n",
       "      <th>detection_rules</th>\n",
       "      <th>genes</th>\n",
       "      <th>subclusters</th>\n",
       "      <th>NRPSs/PKSs</th>\n",
       "      <th>signature_genes</th>\n",
       "      <th>RiPPs</th>\n",
       "      <th>predicted_structure</th>\n",
       "      <th>monomers</th>\n",
       "      <th>model</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AE000511.1</td>\n",
       "      <td>196914</td>\n",
       "      <td>217910</td>\n",
       "      <td>20997</td>\n",
       "      <td>AE000511.1:196914-217910</td>\n",
       "      <td>cf_fatty_acid</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_185;ctg1_186;ctg1_187;ctg1_188;ctg1_189;c...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_190;ctg1_197;ctg1_204</td>\n",
       "      <td>-</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AE000511.1</td>\n",
       "      <td>364402</td>\n",
       "      <td>371194</td>\n",
       "      <td>6793</td>\n",
       "      <td>AE000511.1:364402-371194</td>\n",
       "      <td>cf_putative</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_347;ctg1_348;ctg1_349;ctg1_350;ctg1_351;c...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_352</td>\n",
       "      <td>-</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AE000511.1</td>\n",
       "      <td>582490</td>\n",
       "      <td>603729</td>\n",
       "      <td>21240</td>\n",
       "      <td>AE000511.1:582490-603729</td>\n",
       "      <td>cf_fatty_acid</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_544;ctg1_545;ctg1_546;ctg1_547;ctg1_548;c...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_556;ctg1_557;ctg1_558;ctg1_565</td>\n",
       "      <td>-</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AE000512.1</td>\n",
       "      <td>453877</td>\n",
       "      <td>463748</td>\n",
       "      <td>9872</td>\n",
       "      <td>AE000512.1:453877-463748</td>\n",
       "      <td>cf_putative</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_428;ctg1_429;ctg1_430;ctg1_431;ctg1_432;c...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_436</td>\n",
       "      <td>-</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>AE000512.1</td>\n",
       "      <td>612804</td>\n",
       "      <td>620385</td>\n",
       "      <td>7582</td>\n",
       "      <td>AE000512.1:612804-620385</td>\n",
       "      <td>cf_putative</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_574;ctg1_575;ctg1_576;ctg1_577;ctg1_578;c...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>ctg1_577</td>\n",
       "      <td>-</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   coverage  any_exact  max_covered   contig_id  nucl_start  nucl_end  \\\n",
       "0       0.0      False          0.0  AE000511.1      196914    217910   \n",
       "1       0.0      False          0.0  AE000511.1      364402    371194   \n",
       "2       0.0      False          0.0  AE000511.1      582490    603729   \n",
       "3       0.0      False          0.0  AE000512.1      453877    463748   \n",
       "4       0.0      False          0.0  AE000512.1      612804    620385   \n",
       "\n",
       "   nucl_length              candidate_id        classes detection_rules  \\\n",
       "0        20997  AE000511.1:196914-217910  cf_fatty_acid             NaN   \n",
       "1         6793  AE000511.1:364402-371194    cf_putative             NaN   \n",
       "2        21240  AE000511.1:582490-603729  cf_fatty_acid             NaN   \n",
       "3         9872  AE000512.1:453877-463748    cf_putative             NaN   \n",
       "4         7582  AE000512.1:612804-620385    cf_putative             NaN   \n",
       "\n",
       "                                               genes  subclusters NRPSs/PKSs  \\\n",
       "0  ctg1_185;ctg1_186;ctg1_187;ctg1_188;ctg1_189;c...          NaN        NaN   \n",
       "1  ctg1_347;ctg1_348;ctg1_349;ctg1_350;ctg1_351;c...          NaN        NaN   \n",
       "2  ctg1_544;ctg1_545;ctg1_546;ctg1_547;ctg1_548;c...          NaN        NaN   \n",
       "3  ctg1_428;ctg1_429;ctg1_430;ctg1_431;ctg1_432;c...          NaN        NaN   \n",
       "4  ctg1_574;ctg1_575;ctg1_576;ctg1_577;ctg1_578;c...          NaN        NaN   \n",
       "\n",
       "                       signature_genes RiPPs  predicted_structure  monomers  \\\n",
       "0           ctg1_190;ctg1_197;ctg1_204     -                  NaN       NaN   \n",
       "1                             ctg1_352     -                  NaN       NaN   \n",
       "2  ctg1_556;ctg1_557;ctg1_558;ctg1_565     -                  NaN       NaN   \n",
       "3                             ctg1_436     -                  NaN       NaN   \n",
       "4                             ctg1_577     -                  NaN       NaN   \n",
       "\n",
       "       model  \n",
       "0  antiSMASH  \n",
       "1  antiSMASH  \n",
       "2  antiSMASH  \n",
       "3  antiSMASH  \n",
       "4  antiSMASH  "
      ]
     },
     "execution_count": 989,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cf_as_coverage = pd.read_csv('../data/bacteria/candidates/antismash-clusterfinder/coverage_antismash.csv')\n",
    "cf_as_coverage['model'] = 'antiSMASH'\n",
    "cf_db_coverage = pd.read_csv('../data/bacteria/candidates/antismash-clusterfinder/coverage_deepbgc.csv')\n",
    "cf_db_coverage['model'] = 'DeepBGC'\n",
    "cf_coverage = pd.concat([cf_as_coverage, cf_db_coverage])\n",
    "cf_coverage.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 940,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>coverage</th>\n",
       "      <th>any_exact</th>\n",
       "      <th>max_covered</th>\n",
       "      <th>contig_id</th>\n",
       "      <th>candidate_id</th>\n",
       "      <th>candidate_hash</th>\n",
       "      <th>avg_prediction</th>\n",
       "      <th>nucl_length</th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>num_bio_domains</th>\n",
       "      <th>num_all_domains</th>\n",
       "      <th>num_proteins</th>\n",
       "      <th>protein_ids</th>\n",
       "      <th>pfam_ids</th>\n",
       "      <th>model</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>AE000512.1</td>\n",
       "      <td>AE000512.1(654943-665871)</td>\n",
       "      <td>8afeb1e3be53ecc97724f88a0faaf51a</td>\n",
       "      <td>0.971369</td>\n",
       "      <td>10929</td>\n",
       "      <td>654943</td>\n",
       "      <td>665871</td>\n",
       "      <td>6</td>\n",
       "      <td>27</td>\n",
       "      <td>7</td>\n",
       "      <td>AE000512_615;AE000512_616;AE000512_617;AE00051...</td>\n",
       "      <td>PF00535;PF10111;PF13439;PF13579;PF13692;PF1352...</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.280181</td>\n",
       "      <td>AE000516.2</td>\n",
       "      <td>AE000516.2(103700-117529)</td>\n",
       "      <td>fcf823667b7fbef4a6c7b803165b7b41</td>\n",
       "      <td>0.977752</td>\n",
       "      <td>13830</td>\n",
       "      <td>103700</td>\n",
       "      <td>117529</td>\n",
       "      <td>5</td>\n",
       "      <td>15</td>\n",
       "      <td>8</td>\n",
       "      <td>AE000516_105;AE000516_106;AE000516_107;AE00051...</td>\n",
       "      <td>PF02720;PF02720;PF00823;PF02668;PF10862;PF1319...</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>AE000516.2</td>\n",
       "      <td>AE000516.2(268777-282728)</td>\n",
       "      <td>a85b14dc0a637ed31e25ac57bf78c2a5</td>\n",
       "      <td>0.949831</td>\n",
       "      <td>13952</td>\n",
       "      <td>268777</td>\n",
       "      <td>282728</td>\n",
       "      <td>5</td>\n",
       "      <td>16</td>\n",
       "      <td>10</td>\n",
       "      <td>AE000516_242;AE000516_244;AE000516_245;AE00051...</td>\n",
       "      <td>PF13579;PF13692;PF13439;PF13524;PF00534;PF1127...</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>AE000516.2</td>\n",
       "      <td>AE000516.2(323452-331781)</td>\n",
       "      <td>62e36433ceeb2caaaeab262312897ffb</td>\n",
       "      <td>0.944847</td>\n",
       "      <td>8330</td>\n",
       "      <td>323452</td>\n",
       "      <td>331781</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>AE000516_288;AE000516_289;AE000516_290;AE00051...</td>\n",
       "      <td>PF01896;PF13193;PF00501;PF02771;PF02770;PF0802...</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>False</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>AE000516.2</td>\n",
       "      <td>AE000516.2(625924-644164)</td>\n",
       "      <td>76b6b525a1dd082f5a157d44bd33316b</td>\n",
       "      <td>0.956126</td>\n",
       "      <td>18241</td>\n",
       "      <td>625924</td>\n",
       "      <td>644164</td>\n",
       "      <td>11</td>\n",
       "      <td>30</td>\n",
       "      <td>15</td>\n",
       "      <td>AE000516_564;AE000516_565;AE000516_566;AE00051...</td>\n",
       "      <td>PF08541;PF08545;PF00108;PF08392;PF00195;PF0104...</td>\n",
       "      <td>antiSMASH</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   coverage  any_exact  max_covered   contig_id               candidate_id  \\\n",
       "0       0.0      False     0.000000  AE000512.1  AE000512.1(654943-665871)   \n",
       "1       1.0      False     0.280181  AE000516.2  AE000516.2(103700-117529)   \n",
       "2       0.0      False     0.000000  AE000516.2  AE000516.2(268777-282728)   \n",
       "3       0.0      False     0.000000  AE000516.2  AE000516.2(323452-331781)   \n",
       "4       0.0      False     0.000000  AE000516.2  AE000516.2(625924-644164)   \n",
       "\n",
       "                     candidate_hash  avg_prediction  nucl_length  nucl_start  \\\n",
       "0  8afeb1e3be53ecc97724f88a0faaf51a        0.971369        10929      654943   \n",
       "1  fcf823667b7fbef4a6c7b803165b7b41        0.977752        13830      103700   \n",
       "2  a85b14dc0a637ed31e25ac57bf78c2a5        0.949831        13952      268777   \n",
       "3  62e36433ceeb2caaaeab262312897ffb        0.944847         8330      323452   \n",
       "4  76b6b525a1dd082f5a157d44bd33316b        0.956126        18241      625924   \n",
       "\n",
       "   nucl_end  num_bio_domains  num_all_domains  num_proteins  \\\n",
       "0    665871                6               27             7   \n",
       "1    117529                5               15             8   \n",
       "2    282728                5               16            10   \n",
       "3    331781                6               12             7   \n",
       "4    644164               11               30            15   \n",
       "\n",
       "                                         protein_ids  \\\n",
       "0  AE000512_615;AE000512_616;AE000512_617;AE00051...   \n",
       "1  AE000516_105;AE000516_106;AE000516_107;AE00051...   \n",
       "2  AE000516_242;AE000516_244;AE000516_245;AE00051...   \n",
       "3  AE000516_288;AE000516_289;AE000516_290;AE00051...   \n",
       "4  AE000516_564;AE000516_565;AE000516_566;AE00051...   \n",
       "\n",
       "                                            pfam_ids      model  \n",
       "0  PF00535;PF10111;PF13439;PF13579;PF13692;PF1352...  antiSMASH  \n",
       "1  PF02720;PF02720;PF00823;PF02668;PF10862;PF1319...  antiSMASH  \n",
       "2  PF13579;PF13692;PF13439;PF13524;PF00534;PF1127...  antiSMASH  \n",
       "3  PF01896;PF13193;PF00501;PF02771;PF02770;PF0802...  antiSMASH  \n",
       "4  PF08541;PF08545;PF00108;PF08392;PF00195;PF0104...  antiSMASH  "
      ]
     },
     "execution_count": 940,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "db_as_coverage = pd.read_csv('../data/bacteria/candidates/{}/coverage_antismash.csv'.format(DEEPBGC_PATH))\n",
    "db_as_coverage['model'] = 'antiSMASH'\n",
    "db_cf_coverage = pd.read_csv('../data/bacteria/candidates/{}/coverage_clusterfinder.csv'.format(DEEPBGC_PATH))\n",
    "db_cf_coverage['model'] = 'ClusterFinder'\n",
    "db_coverage = pd.concat([db_as_coverage, db_cf_coverage])\n",
    "db_coverage.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 955,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x144 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, 2, figsize=(7, 2))\n",
    "plot_coverage_distribution(cf_coverage, ax=axes[0])\n",
    "plot_coverage_distribution(db_coverage, ax=axes[1]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Selected candidates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 697,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from Bio import SeqIO\n",
    "from Bio.SeqRecord import SeqRecord\n",
    "from Bio.SeqFeature import FeatureLocation\n",
    "import matplotlib.transforms as transforms\n",
    "from dna_features_viewer import BiopythonTranslator, GraphicRecord, GraphicFeature\n",
    "\n",
    "result_path = '../data/bacteria/candidates/{}/novel.gbk'.format(DEEPBGC_PATH)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 698,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "Credits: DNA features viewer\n",
    "From: https://github.com/Edinburgh-Genome-Foundry/DnaFeaturesViewer\n",
    "\"\"\"\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.patches as mpatches\n",
    "from matplotlib.ticker import MaxNLocator\n",
    "\n",
    "from dna_features_viewer.tools import (change_luminosity, get_text_box, compute_features_levels,\n",
    "                    bokeh_feature_patch)\n",
    "from dna_features_viewer.biotools import extract_translation\n",
    "from dna_features_viewer.GraphicFeature import GraphicFeature\n",
    "\n",
    "from Bio.Seq import Seq\n",
    "from Bio.SeqRecord import SeqRecord\n",
    "from Bio.SeqFeature import FeatureLocation, SeqFeature\n",
    "from Bio.Alphabet import DNAAlphabet\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "class CustomGraphicRecord(GraphicRecord):\n",
    "\n",
    "    def __init__(self, **kwargs):\n",
    "        super().__init__(**kwargs)\n",
    "\n",
    "    def plot_feature(self, ax, feature, level, linewidth=1.0):\n",
    "        x1, x2 = feature.start, feature.end\n",
    "        if feature.open_left:\n",
    "            x1 -= 1\n",
    "        if feature.open_right:\n",
    "            x2 += 1\n",
    "        if feature.strand == -1:\n",
    "            x1, x2 = x2, x1\n",
    "        x1, x2 = x1 - 0.5, x2 - 0.5\n",
    "        is_undirected = feature.strand not in (-1, 1)\n",
    "        head_is_cut = ((feature.strand == 1 and feature.open_right) or\n",
    "                       (feature.strand == -1 and feature.open_left))\n",
    "        head_length = (0.001 if (is_undirected or head_is_cut) else\n",
    "                       max(0.3 * feature.thickness, 5))\n",
    "        \n",
    "        arrowstyle = mpatches.ArrowStyle.Simple(head_width=feature.thickness,\n",
    "                                                tail_width=feature.thickness,\n",
    "                                                head_length=head_length)\n",
    "        y = self.feature_level_width * level\n",
    "        patch = mpatches.FancyArrowPatch([x1, y], [x2, y],\n",
    "                                         shrinkA=0.0, shrinkB=0.0,\n",
    "                                         arrowstyle=arrowstyle,\n",
    "                                         facecolor=feature.color, zorder=0,\n",
    "                                         linewidth=feature.linewidth)\n",
    "        ax.add_patch(patch)\n",
    "        return patch\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 699,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "contig_id\n",
       "000749207dc6a8d25a7b67282a2fe375    Saccharide\n",
       "00244b39ef89b1f18ccf4da018657383    Saccharide\n",
       "0024a1c619f0dba6e8420c60b6c815e7           NRP\n",
       "002d6748dad4145b1f09ca9dc93ead44    Saccharide\n",
       "003041bd79286a337bdd94c332198c33    Saccharide\n",
       "Name: classes, dtype: object"
      ]
     },
     "execution_count": 699,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cand_classes = pd.read_csv('../data/bacteria/candidates/{}/classes/class-domain_set-rf/classes.csv'.format(DEEPBGC_PATH)).set_index('contig_id')['classes']\n",
    "cand_classes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 700,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Accession\n",
       "PF00109    PKS\n",
       "PF02801    PKS\n",
       "PF08659    PKS\n",
       "PF00378    PKS\n",
       "PF08541    PKS\n",
       "Name: Type, dtype: object"
      ]
     },
     "execution_count": 700,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pfam_classes = pd.read_csv('../data/clusterfinder/supplement/pfam_classes.tsv', sep='\\t')\n",
    "pfam_classes = pfam_classes.drop_duplicates(subset=['Accession'], keep='first')\n",
    "pfam_classes = pfam_classes.set_index('Accession')['Type']\n",
    "pfam_classes.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 701,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "PFAM_COLORS = {\n",
    "    'PKS':'darkred',\n",
    "    'NRPS':'darkred',\n",
    "    'Terpene':'darkred',\n",
    "    'Oligosaccharide':'darkred',\n",
    "    'Special':'darkred',\n",
    "    'Ectoine':'darkred',\n",
    "    'Siderophore':'darkred',\n",
    "    'Butyrolactone':'darkred',\n",
    "    'Beta-lactam':'darkred',\n",
    "    'Cyclodipeptide':'darkred',\n",
    "    'Aminoglycoside':'darkred',\n",
    "    'Indolocarbazole':'darkred',\n",
    "    'Nucleoside':'darkred',\n",
    "    'Phosphoglycolipid':'darkred',\n",
    "    'Redox':'blueviolet',\n",
    "    'Group_Transfer':'orange',\n",
    "    'Regulatory':'green',\n",
    "    'Transport':'blue',\n",
    "    'Ribpept':'darkred',\n",
    "    'Other':'grey'\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 702,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def group_genes(domains):\n",
    "    first = domains.iloc[0]\n",
    "    classes = pfam_classes.reindex(domains['pfam_id'].unique()).dropna().unique()\n",
    "    if not len(classes):\n",
    "        classes = ['Other']\n",
    "    return pd.Series(dict(\n",
    "        start=first['gene_start'], \n",
    "        end=first['gene_end'],\n",
    "        strand=first['gene_strand'],\n",
    "        pfam_ids=','.join(domains['pfam_id']),\n",
    "        classes=','.join(classes),\n",
    "        first_class=classes[0]\n",
    "    ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 703,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_domain_genes(cand_domains, figure_width=10, **kwargs):\n",
    "    cand_genes = cand_domains.groupby('protein_id').apply(group_genes)\n",
    "    features = cand_genes.apply(lambda row: GraphicFeature(\n",
    "        start=int(row['start']), \n",
    "        end=int(row['end']), \n",
    "        strand=int(row['strand']), \n",
    "        label='',\n",
    "        color=PFAM_COLORS.get(row['first_class'], 'white')),\n",
    "        axis=1\n",
    "    )\n",
    "    seq_len = cand_domains['gene_end'].max()\n",
    "    graphic_record = CustomGraphicRecord(sequence_length=seq_len, features=features)\n",
    "    graphic_record.feature_level_width = 0\n",
    "    graphic_record.features\n",
    "    ax, _ = graphic_record.plot(figure_width=figure_width, **kwargs)\n",
    "\n",
    "    patches = []\n",
    "    #patches.append(mpatches.Patch(facecolor='darkred', label='Class-specific', edgecolor='black', linewidth=1))\n",
    "    for label, color in PFAM_COLORS.items():\n",
    "        if label not in cand_genes['first_class'].values:\n",
    "            continue\n",
    "        patches.append(mpatches.Patch(facecolor=color, label=label, edgecolor='black', linewidth=1))\n",
    "    \n",
    "    ax.legend(handles=patches, loc='best', bbox_to_anchor=(1.1, 1), frameon=False)\n",
    "    return ax.figure\n",
    "\n",
    "def plot_genes(candidate, **kwargs):\n",
    "    cand_domains = pd.read_csv('../data/bacteria/domains/{}.domains.csv'.format(candidate['contig_id']))\n",
    "    cand_domains = cand_domains[cand_domains['gene_start'] >= candidate['nucl_start']]\n",
    "    cand_domains = cand_domains[cand_domains['gene_end'] <= candidate['nucl_end']]\n",
    "    cand_domains['gene_start'] -= candidate['nucl_start']\n",
    "    cand_domains['gene_end'] -= candidate['nucl_start']\n",
    "    return plot_domain_genes(cand_domains, **kwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 704,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "PALETTE = {\n",
    "    'MIBiG Polyketide' : 'deeppink',\n",
    "    'MIBiG NRP' : 'lime',\n",
    "    'MIBiG RiPP' : 'blue',\n",
    "    'MIBiG Other' : 'darkcyan',\n",
    "    'MIBiG Saccharide' : 'orange',\n",
    "    'MIBiG Terpene': 'turquoise',\n",
    "    'MIBiG Alkaloid' : 'purple',\n",
    "    'MIBiG ?' : None,\n",
    "    'Predicted Polyketide' : 'deeppink:P',\n",
    "    'Predicted NRP' : 'lime:P',\n",
    "    'Predicted RiPP' : 'blue:P',\n",
    "    'Predicted Other' : 'darkcyan:P',\n",
    "    'Predicted Saccharide' : 'orange:P',\n",
    "    'Predicted Terpene': 'turquoise:P',\n",
    "    'Predicted Alkaloid' : 'purple:P'\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 705,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "227"
      ]
     },
     "execution_count": 705,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(novel[novel['num_all_domains'] >= 5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 706,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "227\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>candidate_hash</th>\n",
       "      <th>Alkaloid</th>\n",
       "      <th>NRP</th>\n",
       "      <th>Other</th>\n",
       "      <th>Polyketide</th>\n",
       "      <th>RiPP</th>\n",
       "      <th>Saccharide</th>\n",
       "      <th>Terpene</th>\n",
       "      <th>classes</th>\n",
       "      <th>Antibacterial</th>\n",
       "      <th>...</th>\n",
       "      <th>candidate_id</th>\n",
       "      <th>avg_prediction</th>\n",
       "      <th>nucl_length</th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>num_bio_domains</th>\n",
       "      <th>num_all_domains</th>\n",
       "      <th>num_proteins</th>\n",
       "      <th>protein_ids</th>\n",
       "      <th>pfam_ids</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0289639a11eb66fdf05558583f9a45a3</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>0.194961</td>\n",
       "      <td>0.259762</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.259331</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.360000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP004886.1(1996131-2060701)</td>\n",
       "      <td>0.952772</td>\n",
       "      <td>64571</td>\n",
       "      <td>1996131</td>\n",
       "      <td>2060701</td>\n",
       "      <td>1</td>\n",
       "      <td>77</td>\n",
       "      <td>53</td>\n",
       "      <td>CP004886_2410;CP004886_2411;CP004886_2412;CP00...</td>\n",
       "      <td>PF03007;PF03007;PF06974;PF03007;PF16817;PF0190...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5a2182c16f4ab87f299bf25f31225f87</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.208003</td>\n",
       "      <td>0.305182</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.241884</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.390000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003494.1(1693261-1753880)</td>\n",
       "      <td>0.961767</td>\n",
       "      <td>60620</td>\n",
       "      <td>1693261</td>\n",
       "      <td>1753880</td>\n",
       "      <td>1</td>\n",
       "      <td>70</td>\n",
       "      <td>48</td>\n",
       "      <td>CP003494_1606;CP003494_1607;CP003494_1608;CP00...</td>\n",
       "      <td>PF00934;PF03007;PF06974;PF16817;PF01906;PF0184...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>f3f2b5cb55121b1bea879afa884e3ab3</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>0.030000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.91</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>Saccharide</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>...</td>\n",
       "      <td>CP008802.1(946248-984397)</td>\n",
       "      <td>0.985270</td>\n",
       "      <td>38150</td>\n",
       "      <td>946248</td>\n",
       "      <td>984397</td>\n",
       "      <td>15</td>\n",
       "      <td>90</td>\n",
       "      <td>25</td>\n",
       "      <td>CP008802_802;CP008802_803;CP008802_804;CP00880...</td>\n",
       "      <td>PF12804;PF00483;PF04321;PF01370;PF13460;PF0107...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>f1ca8035648e3be253f1e5ee8e52459c</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.074303</td>\n",
       "      <td>0.350000</td>\n",
       "      <td>0.095697</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.170000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.403500</td>\n",
       "      <td>...</td>\n",
       "      <td>CP006936.1(3460480-3493402)</td>\n",
       "      <td>0.963722</td>\n",
       "      <td>32923</td>\n",
       "      <td>3460480</td>\n",
       "      <td>3493402</td>\n",
       "      <td>2</td>\n",
       "      <td>32</td>\n",
       "      <td>25</td>\n",
       "      <td>CP006936_3288;CP006936_3289;CP006936_3291;CP00...</td>\n",
       "      <td>PF02405;PF12680;PF07366;PF16859;PF00440;PF1397...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>53d05d64e99c1b41d78210f57b6201e6</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.245500</td>\n",
       "      <td>0.129833</td>\n",
       "      <td>0.544667</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.040000</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003233.1(2146676-2176897)</td>\n",
       "      <td>0.957012</td>\n",
       "      <td>30222</td>\n",
       "      <td>2146676</td>\n",
       "      <td>2176897</td>\n",
       "      <td>8</td>\n",
       "      <td>35</td>\n",
       "      <td>23</td>\n",
       "      <td>CP003233_2275;CP003233_2278;CP003233_2279;CP00...</td>\n",
       "      <td>PF01266;PF00890;PF14081;PF00141;PF01475;PF0197...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     candidate_hash  Alkaloid       NRP     Other  Polyketide  \\\n",
       "0  0289639a11eb66fdf05558583f9a45a3      0.04  0.040000  0.194961    0.259762   \n",
       "1  5a2182c16f4ab87f299bf25f31225f87      0.01  0.010000  0.208003    0.305182   \n",
       "2  f3f2b5cb55121b1bea879afa884e3ab3      0.00  0.010000  0.040000    0.030000   \n",
       "3  f1ca8035648e3be253f1e5ee8e52459c      0.00  0.074303  0.350000    0.095697   \n",
       "4  53d05d64e99c1b41d78210f57b6201e6      0.02  0.245500  0.129833    0.544667   \n",
       "\n",
       "   RiPP  Saccharide   Terpene     classes  Antibacterial  \\\n",
       "0  0.30        0.04  0.259331         NaN       0.360000   \n",
       "1  0.33        0.01  0.241884         NaN       0.390000   \n",
       "2  0.01        0.91  0.000000  Saccharide       0.600000   \n",
       "3  0.29        0.04  0.170000         NaN       0.403500   \n",
       "4  0.06        0.06  0.040000  Polyketide       0.333333   \n",
       "\n",
       "                         ...                          \\\n",
       "0                        ...                           \n",
       "1                        ...                           \n",
       "2                        ...                           \n",
       "3                        ...                           \n",
       "4                        ...                           \n",
       "\n",
       "                  candidate_id  avg_prediction  nucl_length nucl_start  \\\n",
       "0  CP004886.1(1996131-2060701)        0.952772        64571    1996131   \n",
       "1  CP003494.1(1693261-1753880)        0.961767        60620    1693261   \n",
       "2    CP008802.1(946248-984397)        0.985270        38150     946248   \n",
       "3  CP006936.1(3460480-3493402)        0.963722        32923    3460480   \n",
       "4  CP003233.1(2146676-2176897)        0.957012        30222    2146676   \n",
       "\n",
       "  nucl_end  num_bio_domains num_all_domains num_proteins  \\\n",
       "0  2060701                1              77           53   \n",
       "1  1753880                1              70           48   \n",
       "2   984397               15              90           25   \n",
       "3  3493402                2              32           25   \n",
       "4  2176897                8              35           23   \n",
       "\n",
       "                                         protein_ids  \\\n",
       "0  CP004886_2410;CP004886_2411;CP004886_2412;CP00...   \n",
       "1  CP003494_1606;CP003494_1607;CP003494_1608;CP00...   \n",
       "2  CP008802_802;CP008802_803;CP008802_804;CP00880...   \n",
       "3  CP006936_3288;CP006936_3289;CP006936_3291;CP00...   \n",
       "4  CP003233_2275;CP003233_2278;CP003233_2279;CP00...   \n",
       "\n",
       "                                            pfam_ids  \n",
       "0  PF03007;PF03007;PF06974;PF03007;PF16817;PF0190...  \n",
       "1  PF00934;PF03007;PF06974;PF16817;PF01906;PF0184...  \n",
       "2  PF12804;PF00483;PF04321;PF01370;PF13460;PF0107...  \n",
       "3  PF02405;PF12680;PF07366;PF16859;PF00440;PF1397...  \n",
       "4  PF01266;PF00890;PF14081;PF00141;PF01475;PF0197...  \n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 706,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novel_long = novel[novel['num_all_domains'] >= 5]\n",
    "print(len(novel_long))\n",
    "novel_long.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#\n",
    "\n",
    "#\n",
    "\n",
    "#\n",
    "\n",
    "#\n",
    "\n",
    "#\n",
    "\n",
    "# TODO move tsne to fpr2\n",
    "\n",
    "#\n",
    "\n",
    "#\n",
    "\n",
    "#\n",
    "\n",
    "#\n",
    "\n",
    "#"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 708,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>a</th>\n",
       "      <th>b</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contig_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BGC0000001.1</th>\n",
       "      <td>-22.674100</td>\n",
       "      <td>-13.696586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000002.1</th>\n",
       "      <td>-34.689125</td>\n",
       "      <td>-13.690722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000003.1</th>\n",
       "      <td>-7.711694</td>\n",
       "      <td>-19.307611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000004.1</th>\n",
       "      <td>-26.408335</td>\n",
       "      <td>-0.806947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000005.1</th>\n",
       "      <td>11.892709</td>\n",
       "      <td>-31.438383</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      a          b\n",
       "contig_id                         \n",
       "BGC0000001.1 -22.674100 -13.696586\n",
       "BGC0000002.1 -34.689125 -13.690722\n",
       "BGC0000003.1  -7.711694 -19.307611\n",
       "BGC0000004.1 -26.408335  -0.806947\n",
       "BGC0000005.1  11.892709 -31.438383"
      ]
     },
     "execution_count": 708,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tsne = pd.read_csv('../data/bacteria/candidates/novel-{}/tsne/mibig_bgcs_all.pfam2vec-mean.tsne2.csv'.format(DEEPBGC_PATH)).set_index('contig_id')\n",
    "#tsne = pd.read_csv('../data/bacteria/candidates/novel-{}/tsne/mibig_bgcs_all.domain_set.tsne2.csv'.format(DEEPBGC_PATH)).set_index('contig_id')\n",
    "tsne.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 709,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>a</th>\n",
       "      <th>b</th>\n",
       "      <th>marker</th>\n",
       "      <th>alpha</th>\n",
       "      <th>size</th>\n",
       "      <th>group</th>\n",
       "      <th>color</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contig_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0289639a11eb66fdf05558583f9a45a3</th>\n",
       "      <td>-49.043210</td>\n",
       "      <td>13.077333</td>\n",
       "      <td>P</td>\n",
       "      <td>0.8</td>\n",
       "      <td>30</td>\n",
       "      <td>Predicted (no class)</td>\n",
       "      <td>grey</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5a2182c16f4ab87f299bf25f31225f87</th>\n",
       "      <td>-49.218002</td>\n",
       "      <td>13.114355</td>\n",
       "      <td>P</td>\n",
       "      <td>0.8</td>\n",
       "      <td>30</td>\n",
       "      <td>Predicted (no class)</td>\n",
       "      <td>grey</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f3f2b5cb55121b1bea879afa884e3ab3</th>\n",
       "      <td>21.367891</td>\n",
       "      <td>45.199375</td>\n",
       "      <td>P</td>\n",
       "      <td>0.8</td>\n",
       "      <td>30</td>\n",
       "      <td>Predicted Saccharide</td>\n",
       "      <td>orange:P</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f1ca8035648e3be253f1e5ee8e52459c</th>\n",
       "      <td>-34.637905</td>\n",
       "      <td>13.503100</td>\n",
       "      <td>P</td>\n",
       "      <td>0.8</td>\n",
       "      <td>30</td>\n",
       "      <td>Predicted (no class)</td>\n",
       "      <td>grey</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53d05d64e99c1b41d78210f57b6201e6</th>\n",
       "      <td>-24.450462</td>\n",
       "      <td>8.060456</td>\n",
       "      <td>P</td>\n",
       "      <td>0.8</td>\n",
       "      <td>30</td>\n",
       "      <td>Predicted Polyketide</td>\n",
       "      <td>deeppink:P</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          a          b marker  alpha  size  \\\n",
       "contig_id                                                                    \n",
       "0289639a11eb66fdf05558583f9a45a3 -49.043210  13.077333      P    0.8    30   \n",
       "5a2182c16f4ab87f299bf25f31225f87 -49.218002  13.114355      P    0.8    30   \n",
       "f3f2b5cb55121b1bea879afa884e3ab3  21.367891  45.199375      P    0.8    30   \n",
       "f1ca8035648e3be253f1e5ee8e52459c -34.637905  13.503100      P    0.8    30   \n",
       "53d05d64e99c1b41d78210f57b6201e6 -24.450462   8.060456      P    0.8    30   \n",
       "\n",
       "                                                 group       color  \n",
       "contig_id                                                           \n",
       "0289639a11eb66fdf05558583f9a45a3  Predicted (no class)        grey  \n",
       "5a2182c16f4ab87f299bf25f31225f87  Predicted (no class)        grey  \n",
       "f3f2b5cb55121b1bea879afa884e3ab3  Predicted Saccharide    orange:P  \n",
       "f1ca8035648e3be253f1e5ee8e52459c  Predicted (no class)        grey  \n",
       "53d05d64e99c1b41d78210f57b6201e6  Predicted Polyketide  deeppink:P  "
      ]
     },
     "execution_count": 709,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "novel_samples = tsne.loc[novel_long['candidate_hash'].unique()].copy()\n",
    "prefix = 'Predicted '\n",
    "novel_samples['marker'] = 'P'\n",
    "novel_samples['alpha'] = 0.8\n",
    "novel_samples['size'] = 30\n",
    "novel_samples['group'] = cand_classes.loc[novel_samples.index].fillna('(no class)').apply(lambda c: prefix+c.replace(';', ';'+prefix))\n",
    "novel_samples['color'] = novel_samples['group'].apply(lambda g: apply_group_color(g, PALETTE))\n",
    "novel_samples.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 710,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>a</th>\n",
       "      <th>b</th>\n",
       "      <th>marker</th>\n",
       "      <th>alpha</th>\n",
       "      <th>size</th>\n",
       "      <th>group</th>\n",
       "      <th>color</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contig_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BGC0000001.1</th>\n",
       "      <td>-22.674100</td>\n",
       "      <td>-13.696586</td>\n",
       "      <td>o</td>\n",
       "      <td>0.15</td>\n",
       "      <td>20</td>\n",
       "      <td>MIBiG Polyketide</td>\n",
       "      <td>deeppink</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000002.1</th>\n",
       "      <td>-34.689125</td>\n",
       "      <td>-13.690722</td>\n",
       "      <td>o</td>\n",
       "      <td>0.15</td>\n",
       "      <td>20</td>\n",
       "      <td>MIBiG Polyketide</td>\n",
       "      <td>deeppink</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000003.1</th>\n",
       "      <td>-7.711694</td>\n",
       "      <td>-19.307611</td>\n",
       "      <td>o</td>\n",
       "      <td>0.15</td>\n",
       "      <td>20</td>\n",
       "      <td>MIBiG Polyketide</td>\n",
       "      <td>deeppink</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000004.1</th>\n",
       "      <td>-26.408335</td>\n",
       "      <td>-0.806947</td>\n",
       "      <td>o</td>\n",
       "      <td>0.15</td>\n",
       "      <td>20</td>\n",
       "      <td>MIBiG Polyketide</td>\n",
       "      <td>deeppink</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000005.1</th>\n",
       "      <td>11.892709</td>\n",
       "      <td>-31.438383</td>\n",
       "      <td>o</td>\n",
       "      <td>0.15</td>\n",
       "      <td>20</td>\n",
       "      <td>MIBiG Polyketide</td>\n",
       "      <td>deeppink</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      a          b marker  alpha  size             group  \\\n",
       "contig_id                                                                  \n",
       "BGC0000001.1 -22.674100 -13.696586      o   0.15    20  MIBiG Polyketide   \n",
       "BGC0000002.1 -34.689125 -13.690722      o   0.15    20  MIBiG Polyketide   \n",
       "BGC0000003.1  -7.711694 -19.307611      o   0.15    20  MIBiG Polyketide   \n",
       "BGC0000004.1 -26.408335  -0.806947      o   0.15    20  MIBiG Polyketide   \n",
       "BGC0000005.1  11.892709 -31.438383      o   0.15    20  MIBiG Polyketide   \n",
       "\n",
       "                 color  \n",
       "contig_id               \n",
       "BGC0000001.1  deeppink  \n",
       "BGC0000002.1  deeppink  \n",
       "BGC0000003.1  deeppink  \n",
       "BGC0000004.1  deeppink  \n",
       "BGC0000005.1  deeppink  "
      ]
     },
     "execution_count": 710,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mibig_samples = tsne.loc[mibig.index].dropna().copy()\n",
    "prefix = 'MIBiG '\n",
    "mibig_samples['marker'] = 'o'\n",
    "mibig_samples['alpha'] = 0.15\n",
    "mibig_samples['size'] = 20\n",
    "mibig_samples['group'] = mibig.loc[mibig_samples.index]['classes'].apply(lambda c: prefix+c.replace(';', ';'+prefix) if isinstance(c, str) else c)\n",
    "mibig_samples['color'] = mibig_samples['group'].apply(lambda g: apply_group_color(g, PALETTE))\n",
    "mibig_samples.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 711,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "first_cand = novel[novel['candidate_hash']=='ffd47679ea075ec67ed6bb995467cbcb'].iloc[0]\n",
    "second_cand = novel[novel['candidate_hash']=='307e7e7ef2b7256aa60260c3f20ed974'].iloc[0]\n",
    "third_cand = novel[novel['candidate_hash']=='ae833e7695e2a30dbf51c4db523a46b5'].iloc[0]\n",
    "first_tsne = tsne.loc[first_cand['candidate_hash']]\n",
    "second_tsne = tsne.loc[second_cand['candidate_hash']]\n",
    "third_tsne = tsne.loc[third_cand['candidate_hash']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 712,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Numeric:  ['pfam2vec mean: tsne1', 'pfam2vec mean: tsne2', 'alpha', 'size']\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 324x324 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "samples = pd.concat([mibig_samples, novel_samples])\n",
    "samples = samples.rename(columns={'a': 'pfam2vec mean: tsne1', 'b': 'pfam2vec mean: tsne2'})\n",
    "fig = group_scatter_plot(samples=samples, groupby='group', figure_size=4.5, verbose=0, legend_hybrids=False, legend_loc='outside best', linewidth=0.8)\n",
    "fig.axes[0].plot(first_tsne['a'], first_tsne['b'], marker='o', markersize=20, markeredgecolor='black', markerfacecolor=(0,0,0,0), lw=0.5)\n",
    "#fig.axes[0].plot(third_tsne['a'], third_tsne['b'], marker='o', markersize=20, markeredgecolor='red', markerfacecolor=(0,0,0,0), lw=0.5)\n",
    "#fig.axes[0].set_xlim([None, 115])\n",
    "#fig.axes[0].set_ylim([-65, 60])\n",
    "fig.axes[0].set_xticks([-50,-25,0,25,50])\n",
    "fig.axes[0].set_yticks([-50,-25,0,25,50])\n",
    "fig.axes[0].get_legend().remove()\n",
    "fig.savefig('../data/figures/paper/main/5a.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 713,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x134.64 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_genes(novel[novel['candidate_hash']=='ffd47679ea075ec67ed6bb995467cbcb'].iloc[0], figure_width=5)\n",
    "#fig.savefig('../data/figures/paper/main/5b.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1001,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>model</th>\n",
       "      <th>candidate_hash</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38808</th>\n",
       "      <td>290351</td>\n",
       "      <td>306057</td>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6482</th>\n",
       "      <td>314916</td>\n",
       "      <td>333380</td>\n",
       "      <td>DeepBGC</td>\n",
       "      <td>ffd47679ea075ec67ed6bb995467cbcb</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38809</th>\n",
       "      <td>390102</td>\n",
       "      <td>404021</td>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       nucl_start  nucl_end          model                    candidate_hash\n",
       "38808      290351    306057  ClusterFinder                               NaN\n",
       "6482       314916    333380        DeepBGC  ffd47679ea075ec67ed6bb995467cbcb\n",
       "38809      390102    404021  ClusterFinder                               NaN"
      ]
     },
     "execution_count": 1001,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neighbors = candidates[candidates['contig_id'] == 'CP004886.1']\n",
    "neighbors = neighbors[(neighbors['nucl_start'] > 270000) & (neighbors['nucl_end'] < 450000)]\n",
    "neighbors = neighbors[['nucl_start','nucl_end','model','candidate_hash']].sort_values(by='nucl_start')\n",
    "neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 999,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8859"
      ]
     },
     "execution_count": 999,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "314916-306057"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 998,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "56722"
      ]
     },
     "execution_count": 998,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "390102-333380"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 562,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CP004886.1 of length 404021\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x54 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_candidate_region('ffd47679ea075ec67ed6bb995467cbcb', start=0, end=450000)\n",
    "fig.set_figwidth(4)\n",
    "fig.set_figheight(0.75)\n",
    "#fig.savefig('../data/figures/paper/main/5c.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x134.64 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_genes(novel[novel['candidate_hash']=='307e7e7ef2b7256aa60260c3f20ed974'].iloc[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 563,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CP003234.1 of length 403415\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x54 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_candidate_region('307e7e7ef2b7256aa60260c3f20ed974', start=0, end=450000)\n",
    "fig.set_figwidth(4)\n",
    "fig.set_figheight(0.75)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x134.64 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "candidate = novel[novel['candidate_hash']=='ae833e7695e2a30dbf51c4db523a46b5'].iloc[0]\n",
    "plot_genes(candidate);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CP002299.1 of length 4442946\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x54 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_candidate_region('ae833e7695e2a30dbf51c4db523a46b5', start=3800000, end=4400000)\n",
    "fig.set_figwidth(4)\n",
    "fig.set_figheight(0.75)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Not used, kept for reference\n",
    "def plot_record_domains(record):\n",
    "    translator = BiopythonTranslator()\n",
    "    graphic_record = CustomGraphicRecord(sequence=record.seq, features=[translator.translate_feature(f) for f in record.features if f.type == 'CDS'])\n",
    "    graphic_record.feature_level_width = 0\n",
    "    for feature in graphic_record.features:\n",
    "        feature.label = ''\n",
    "        feature.thickness = 20\n",
    "        feature.linewidth=1\n",
    "    ax, _ = graphic_record.plot(figure_width=10, annotate_inline=True)\n",
    "    graphic_record = CustomGraphicRecord(sequence=record.seq, features=[translator.translate_feature(f) for f in record.features if f.type == 'PFAM_domain'])\n",
    "    graphic_record.feature_level_width = 2\n",
    "    for feature in graphic_record.features:\n",
    "        feature.label = ''\n",
    "        feature.thickness = 13\n",
    "        feature.color = 'orange'\n",
    "        feature.linewidth=1\n",
    "    ax, _ = graphic_record.plot(annotate_inline=True, ax=ax, level_offset=2)\n",
    "    ax.set_ylim([-5, None])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Similar to selected candidate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 714,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics.pairwise import cosine_similarity, cosine_distances"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 715,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>pfam_id</th>\n",
       "      <th>PF00004</th>\n",
       "      <th>PF00005</th>\n",
       "      <th>PF00006</th>\n",
       "      <th>PF00009</th>\n",
       "      <th>PF00011</th>\n",
       "      <th>PF00012</th>\n",
       "      <th>PF00013</th>\n",
       "      <th>PF00015</th>\n",
       "      <th>PF00016</th>\n",
       "      <th>PF00018</th>\n",
       "      <th>...</th>\n",
       "      <th>PF17409</th>\n",
       "      <th>PF17410</th>\n",
       "      <th>PF17418</th>\n",
       "      <th>PF17425</th>\n",
       "      <th>PF17456</th>\n",
       "      <th>PF17479</th>\n",
       "      <th>PF17508</th>\n",
       "      <th>PF17590</th>\n",
       "      <th>PF17618</th>\n",
       "      <th>PF17619</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contig_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>000749207dc6a8d25a7b67282a2fe375</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>00244b39ef89b1f18ccf4da018657383</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0024a1c619f0dba6e8420c60b6c815e7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>002d6748dad4145b1f09ca9dc93ead44</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>003041bd79286a337bdd94c332198c33</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 4416 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "pfam_id                           PF00004  PF00005  PF00006  PF00009  PF00011  \\\n",
       "contig_id                                                                       \n",
       "000749207dc6a8d25a7b67282a2fe375        0        0        0        0        0   \n",
       "00244b39ef89b1f18ccf4da018657383        0        0        0        0        0   \n",
       "0024a1c619f0dba6e8420c60b6c815e7        0        0        0        0        0   \n",
       "002d6748dad4145b1f09ca9dc93ead44        0        0        0        0        0   \n",
       "003041bd79286a337bdd94c332198c33        0        0        0        0        0   \n",
       "\n",
       "pfam_id                           PF00012  PF00013  PF00015  PF00016  PF00018  \\\n",
       "contig_id                                                                       \n",
       "000749207dc6a8d25a7b67282a2fe375        0        0        0        0        0   \n",
       "00244b39ef89b1f18ccf4da018657383        0        0        0        0        0   \n",
       "0024a1c619f0dba6e8420c60b6c815e7        0        0        0        0        0   \n",
       "002d6748dad4145b1f09ca9dc93ead44        0        0        0        0        0   \n",
       "003041bd79286a337bdd94c332198c33        0        0        0        0        0   \n",
       "\n",
       "pfam_id                            ...     PF17409  PF17410  PF17418  PF17425  \\\n",
       "contig_id                          ...                                          \n",
       "000749207dc6a8d25a7b67282a2fe375   ...           0        0        0        0   \n",
       "00244b39ef89b1f18ccf4da018657383   ...           0        0        0        0   \n",
       "0024a1c619f0dba6e8420c60b6c815e7   ...           0        0        0        0   \n",
       "002d6748dad4145b1f09ca9dc93ead44   ...           0        0        0        0   \n",
       "003041bd79286a337bdd94c332198c33   ...           0        0        0        0   \n",
       "\n",
       "pfam_id                           PF17456  PF17479  PF17508  PF17590  PF17618  \\\n",
       "contig_id                                                                       \n",
       "000749207dc6a8d25a7b67282a2fe375        0        0        0        0        0   \n",
       "00244b39ef89b1f18ccf4da018657383        0        0        0        0        0   \n",
       "0024a1c619f0dba6e8420c60b6c815e7        0        0        0        0        0   \n",
       "002d6748dad4145b1f09ca9dc93ead44        0        0        0        0        0   \n",
       "003041bd79286a337bdd94c332198c33        0        0        0        0        0   \n",
       "\n",
       "pfam_id                           PF17619  \n",
       "contig_id                                  \n",
       "000749207dc6a8d25a7b67282a2fe375        0  \n",
       "00244b39ef89b1f18ccf4da018657383        0  \n",
       "0024a1c619f0dba6e8420c60b6c815e7        0  \n",
       "002d6748dad4145b1f09ca9dc93ead44        0  \n",
       "003041bd79286a337bdd94c332198c33        0  \n",
       "\n",
       "[5 rows x 4416 columns]"
      ]
     },
     "execution_count": 715,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "domain_set = pd.read_pickle('../data/bacteria/candidates/{}/domain_set.pkl'.format(DEEPBGC_PATH))\n",
    "domain_set.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 716,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "23"
      ]
     },
     "execution_count": 716,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_domain_set = domain_set.loc[first_cand['candidate_hash']]\n",
    "sum(first_domain_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 717,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "contig_id\n",
       "ffd47679ea075ec67ed6bb995467cbcb    1.000000\n",
       "307e7e7ef2b7256aa60260c3f20ed974    0.800198\n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44    0.677672\n",
       "f5e0ed012e76c5eda3c862febdee4e83    0.669711\n",
       "257ddd67560ffe89405860edcfeec6b5    0.638915\n",
       "62e36433ceeb2caaaeab262312897ffb    0.628695\n",
       "205ed77242499c881cc0c1c3cdf71f89    0.613006\n",
       "6ba88b1a3182d0183371fb410f9ce219    0.578315\n",
       "6e6077b2b45de59b3bc8174e15ed36e1    0.551677\n",
       "ad6a07c13956c21f4ea44f40fc801d58    0.541736\n",
       "dtype: float64"
      ]
     },
     "execution_count": 717,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "first_similarity = pd.Series(cosine_similarity([first_domain_set], domain_set)[0], domain_set.index).sort_values(ascending=False)\n",
    "first_similarity[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 972,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "candidate_hash                      ad6a07c13956c21f4ea44f40fc801d58\n",
       "Alkaloid                                                           0\n",
       "NRP                                                         0.558429\n",
       "Other                                                       0.173119\n",
       "Polyketide                                                  0.268452\n",
       "RiPP                                                               0\n",
       "Saccharide                                                     0.005\n",
       "Terpene                                                            0\n",
       "classes                                                          NRP\n",
       "Antibacterial                                                    0.7\n",
       "Cytotoxic                                                       0.01\n",
       "Inhibitor                                                       0.07\n",
       "Antifungal                                                      0.05\n",
       "classes1                                               Antibacterial\n",
       "BGC_ID                                                  BGC0000131.1\n",
       "Similarity                                                  0.363636\n",
       "contig_id                                                 CP003323.1\n",
       "species                         Mycobacterium_intracellulare_MOTT-02\n",
       "cf_prediction                                              0.0218797\n",
       "coverage                                                           1\n",
       "any_exact                                                      False\n",
       "max_covered                                                 0.230431\n",
       "candidate_id                             CP003323.1(5024974-5036557)\n",
       "avg_prediction                                              0.948203\n",
       "nucl_length                                                    11584\n",
       "nucl_start                                                   5024974\n",
       "nucl_end                                                     5036557\n",
       "num_bio_domains                                                    7\n",
       "num_all_domains                                                   12\n",
       "num_proteins                                                       7\n",
       "protein_ids        CP003323_4701;CP003323_4702;CP003323_4703;CP00...\n",
       "pfam_ids           PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...\n",
       "Name: 7513, dtype: object"
      ]
     },
     "execution_count": 972,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cand"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 988,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mycobacterium_tuberculosis_str._Haarlem/NITR202: Similarity 100.00%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        0.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.17490265872883534\n",
      "Mycobacterium_tuberculosis_RGTB423: Similarity 80.02%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        0.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.17902029810524653\n",
      "Mycobacterium_intracellulare_MOTT-64: Similarity 67.77%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        1.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.021325113847974652\n",
      "Mycobacterium_avium_subsp._hominissuis_TH135_chromosomal_DNA: Similarity 66.97%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        1.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.5030428732838098\n",
      "Mycobacterium_avium_104: Similarity 63.89%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        1.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.10256551329035243\n",
      "Mycobacterium_tuberculosis_CDC1551: Similarity 62.87%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        0.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.4279500790294473\n",
      "Mycobacterium_indicus_pranii_MTCC_9506: Similarity 61.30%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        1.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.021556083342187497\n",
      "Mycobacterium_sp._MOTT36Y: Similarity 57.83%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        1.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.02298360345919584\n",
      "Mycobacterium_tuberculosis_EAI5/NITR206: Similarity 55.17%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        0.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.5097029665784669\n",
      "Mycobacterium_intracellulare_MOTT-02: Similarity 54.17%:\n",
      "model\n",
      "ClusterFinder    0.0\n",
      "antiSMASH        1.0\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.021879661730219328\n",
      "Mycobacterium_kansasii_ATCC_12478: Similarity 50.57%:\n",
      "model\n",
      "ClusterFinder    0.884869\n",
      "antiSMASH        0.000000\n",
      "Name: coverage, dtype: float64\n",
      "CF prediction:  0.6752302330386372\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x396 with 11 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(11, 1, figsize=(5, 5.5))\n",
    "i = 0\n",
    "for candidate_hash, sim in first_similarity[:11].iteritems():\n",
    "    matching = deepbgc[deepbgc['candidate_hash']==candidate_hash]\n",
    "    cand = matching.iloc[0]\n",
    "    novelty = db_coverage.set_index('candidate_id').loc[matching['candidate_id']].groupby('model')['coverage'].max()\n",
    "    print('{}: Similarity {:.2f}%:'.format(cand['species'], sim*100))\n",
    "    print(novelty)\n",
    "    print('CF prediction: ', matching['cf_prediction'].max())\n",
    "    plot_genes(cand, figure_width=7, ax=axes[i])\n",
    "    axes[i].legend([]).remove()\n",
    "    axes[i].set_xticks([])\n",
    "    i += 1\n",
    "fig.savefig('../data/figures/paper/supplementary/11a.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 719,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>pfam_id</th>\n",
       "      <th>PF00004</th>\n",
       "      <th>PF00005</th>\n",
       "      <th>PF00006</th>\n",
       "      <th>PF00009</th>\n",
       "      <th>PF00010</th>\n",
       "      <th>PF00011</th>\n",
       "      <th>PF00012</th>\n",
       "      <th>PF00013</th>\n",
       "      <th>PF00015</th>\n",
       "      <th>PF00022</th>\n",
       "      <th>...</th>\n",
       "      <th>PF17387</th>\n",
       "      <th>PF17389</th>\n",
       "      <th>PF17390</th>\n",
       "      <th>PF17391</th>\n",
       "      <th>PF17392</th>\n",
       "      <th>PF17414</th>\n",
       "      <th>PF17425</th>\n",
       "      <th>PF17508</th>\n",
       "      <th>PF17584</th>\n",
       "      <th>PF17608</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>contig_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>BGC0000001.1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000002.1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000003.1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000004.1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>BGC0000005.1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3294 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "pfam_id       PF00004  PF00005  PF00006  PF00009  PF00010  PF00011  PF00012  \\\n",
       "contig_id                                                                     \n",
       "BGC0000001.1        0        1        0        0        0        0        0   \n",
       "BGC0000002.1        0        1        0        0        0        0        0   \n",
       "BGC0000003.1        0        0        0        0        0        0        0   \n",
       "BGC0000004.1        0        0        0        0        0        0        0   \n",
       "BGC0000005.1        0        0        0        0        0        0        0   \n",
       "\n",
       "pfam_id       PF00013  PF00015  PF00022   ...     PF17387  PF17389  PF17390  \\\n",
       "contig_id                                 ...                                 \n",
       "BGC0000001.1        0        0        0   ...           0        0        0   \n",
       "BGC0000002.1        0        0        0   ...           0        0        0   \n",
       "BGC0000003.1        0        0        0   ...           0        0        0   \n",
       "BGC0000004.1        0        0        0   ...           0        0        0   \n",
       "BGC0000005.1        0        0        0   ...           0        0        0   \n",
       "\n",
       "pfam_id       PF17391  PF17392  PF17414  PF17425  PF17508  PF17584  PF17608  \n",
       "contig_id                                                                    \n",
       "BGC0000001.1        0        0        0        0        0        0        0  \n",
       "BGC0000002.1        0        0        0        0        0        0        0  \n",
       "BGC0000003.1        0        0        0        0        0        0        0  \n",
       "BGC0000004.1        0        0        0        0        0        0        0  \n",
       "BGC0000005.1        0        0        0        0        0        0        0  \n",
       "\n",
       "[5 rows x 3294 columns]"
      ]
     },
     "execution_count": 719,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mibig_domain_set = pd.read_pickle('../data/training/positive/mibig_bgcs_all.domain_set.pkl')\n",
    "mibig_domain_set.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 720,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "contig_id\n",
       "BGC0000373.1    0.389249\n",
       "BGC0000961.1    0.362738\n",
       "BGC0000131.1    0.353553\n",
       "BGC0000312.1    0.350823\n",
       "BGC0000913.1    0.333712\n",
       "BGC0000328.1    0.322749\n",
       "BGC0000456.4    0.316228\n",
       "BGC0001253.1    0.316228\n",
       "BGC0001180.2    0.316228\n",
       "BGC0000095.1    0.316228\n",
       "dtype: float64"
      ]
     },
     "execution_count": 720,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "common_features = np.intersect1d(domain_set.columns, mibig_domain_set.columns)\n",
    "first_similarity_mibig = pd.Series(cosine_similarity([first_domain_set.loc[common_features]], mibig_domain_set[common_features])[0], mibig_domain_set.index).sort_values(ascending=False)\n",
    "first_similarity_mibig[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 721,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BGC0000373.1: Similarity 38.92%:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x28.8 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BGC0000961.1: Similarity 36.27%:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x28.8 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BGC0000131.1: Similarity 35.36%:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x28.8 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BGC0000312.1: Similarity 35.08%:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x28.8 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "BGC0000913.1: Similarity 33.37%:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZwAAAA1CAYAAAB1JV2rAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAE59JREFUeJztnXuYFNWVwH+nnzM9Iw/lEQIaWCH4CC6DEhQIJiqGGBXkNeQBYnbVlUiyGvNcEzW70dUYY2JIVk3UxIQBQkBE4uoCAuHDlWB4DVFBfCAQURmQmemZru6us39UtTRDz9DT013dnb2/7+vv67r3Vp1z6j7OrXur7hVVxWAwGAyGQuMrtgIGg8Fg+P+BcTgGg8Fg8ATjcAwGg8HgCcbhGAwGg8ETjMMxGAwGgycYh2MwGAwGTzAOx2AwGAyeYByOwWAwGDzBOByDwWAweEKgM4nD4fCD7cUlEomlyWTyv7O9loj0Ish3ESo6o0PesPipqu5oGywiQ4LB4M0i4okzVlUrHo/fCvSIhPmWT7LrBCRtDrdY3KaqrenhIhIIBAL3+Hy+quPOSSY3JRKJh9u7ZjAYvNnn8w3ttBEeYtv2wUQicbuqWpnig8HgTT6f7wyv9So0qpqMx+M/UtXdqbBIWG70+xhWCHnxBGta41qXOhaRESG4vhCy0rFhbwJ+oKq2KzccCAT+0+fzRQotOx1VbYrH47eraqOrhy8Ad/mgh5d65IpCa9y5j+90lC4YDH7V5/Od1RVZsVgs63IhnVnaZvTo0dq9e/fjwpPJJGvWrInG4/HhqrrrhEJFhDArGcw4Tu2c08sL9cTZy02qOq+NXhWhUKi+pqbmH3r27CleqLJu3brWaDQ6Bhg0tB+/nXNJdg54ySZaX3ydxxpb9Ib0cBHp4/f7940fP/64+7p27dqW1tbWybZtZ+wYhMPhhtGjR/esqChOHyAbduzY0fL2228/aFnWTZniw+HwexdccMEplZWVXqtWUA4ePKhbt27daVnWOSln26NK9s65hP4fOr5Kdglb4XuLaWls5SJV/V8AEfnm6XDnxwo8KvICRN+D2+Kq97pyTw0Gg69efPHFoULKbcvu3butN998c2ksFpvh6lEt8P6VZTIq9CYkXoL1MbhI22nkfT7fjOrq6l+NGTMmZ2f+/vvvs2HDhqzbyk419meffTYDBgzIGOf3+ytWr179hIgMV9V4hxfy82W6MYqrCBTB3cA7JNl7fHAoFLpv4MCB/SZMmCAinvgbNm7cGItGowAM6o31lQnZOZwvjqVi6C3M8vtkWdLWYxyI3++Pn3/++cfd2b59+1bOnz+/TkSGttfzqampIVOnolQYNmxY5QMPPHCdiCxX1dWZ0gwfPpyePXt6rVpBUVVpaGg4dc+ePfcA/5oKnzkWzuyff3mn9aJy1i9Y6paVIwADwB5b4Ab3TIjcB98XkZWqugUgEAjEzz//fE8dTk1NTWjevHlX+Hy+Wtu2FwL4IFlo+/PFBRC4H0a+C18BftI2XkQ+EgwGH54xY0akf//cC9DevRka0g7I280bNWqUr3///gODweBdHaUTkbPwcTe1VBXF2bSDz+eb4Pf7Z0+aNCnilbPpCidXw6K5RCqC1IlIn2zOGTRoECNHjoyEw+E6zzxqnqmqqmLq1KmRYDC4SEROKbY+XiEiTJ48OeL3+68TkYsLLW/ySJj6cXqcVMEvCy0rnVOAyVAZgidExNNhtHTC4TC1tbURv9//SxE5rVh65IofuBqq/HCniBwz7Coi/nA4vGTcuHGVXXE2uZA3hyMiTJkyJeL3++eIyIXtpAkTYhkTqKRXviR3HRHpEwgE6qZPn14ZiRStjHeaT50Nc8YT6VZJ1g7koosuCnXr1m2U3++/sdD6FYrBgwdTU1NTHQqFfluujjMXqqqqmDZtWqVXznbebCp6RPhswC+fK7SsdEYAQ6FPCB7wUm5b+vfvz7hx4ypCodASyuTJJp1ewCSoDMMyEflg5CQQCNzau3fvoWPGjPF7rVNeb2J1dTVTpkypDAaDi0Xk+DGNEPfxET7MCEqmkRARCYfDdeedd15k0KBBxVan0/xgOqEBJzMqFODL2aQPBALU1tZW+Xy+u0Xk7ELrVyjGjx8frq6u/oTf77+22Lp4yemnn86IESOqw+HwfAq8s0hVBSy9mUg4wEN4OFkuwDTnKWcGMN4ruZkYO3ZsoHfv3mf6/f6vF1OPXDkPZDB8KAz3A4jIKJ/P981p06ZV+Xze+9C8SxwyZAjDhw/vFg6HH0/vfYpPLsXPNVxFpHTcDfj9/i9369ZtlNeTkvkiFIClN1EV9JO1A+nVqxcTJkyoCIVCx/R8yolgMMiMGTOqfD7fj0WkpN+syzfjx48PVVdXj1E47k3EfHPuIPjuVVSGAszycuesSmAWRAJOQ1m0FsPn8zF9+vSIz+f7WjH1yBUBpkNlAGaJyORgMLh00qRJlcWapy2Ii7v00ktDkUjkM0C9iKwRkXX4WcE0KimFESslBHxVRJ4XkZ/U1tZWBQLFmVBS1WrgIeCOhJ1bA/LRfnDfF4lUV/BnYCkQPtE5I0aMkNNOO21QIBDY6ebRGlU9KRf5xaJPnz5ccsklkVAo9GLKBqB033jIE6mnVDx66vjG5fj7dqMfnXzJqKsMAkZCNapFbTW6d+/O5ZdfXonH9ueLCDATKkOBwOIzzjij71lndekt6C5REIcTj8dpbW0FSLhBipCkuRDSuoQtItrS0lJsPbrMvgZUlY7fDkzDtm2ampqwbTvrc0qRI0eO6Anfivw7xMsyayWgOeaZuGNoBIVCDx6emCNHjhRdh67QCKht242NjRRzl+e8OxxVZdmyZS3JZPIhVf1HVf2kql5InAtYTguH8i0xBwQL+ImqjkkkElMXLlwYdR2k96qINAHXAbcFfLm55Od3wb0raGyOcRZwFXDC5mHdunWJQ4cObbRte4ibR58UkcZc5BeLN954g40bNx6Jx+NnpGwA3i+2XoWmpaWFRYsWRQUOeiHvlvlYTTFe4WgH0hO2Aa/Ae4gUtUd44MAB1q5dG8Vj+/PFIWAxtMRte8z+/ftf3rhxo10sXfLucDZv3qyvv/76/rYf5qnqZmzuYBHNJPMtNXds237CsqzFy5cvL8vHnCNRmHI/0dY4V6vqvmzOeeutt9iwYUNLLBabmvqiu9xINbqJROILqnqg2Pp4hary5JNPtsTj8TqEgveSnt4Cj62j0Uqw0MsJjMPAImiJwz9TxCeceDzOggULoslk8mvF1CNXbOA30Jx0Vh14wbKsiStXrmw9cKA4VSavDufgwYM8/fTTLZZlTWy75AoACX5IA9tYV1o9Bcuy5uzateu97du3F1uVTnP9I7Q2xVictPWJbNK3traycOHCaCKRyNpBlRqpp+hEIvE727ZXFFsfL9m6davu3r37Hcuy5hZa1jvvw+fn0dIcYwrgWYfMBn4LzTbcA2zxSm4mnnnmmVg0Gl2pqr8rph65sgoS78KOBNwFoKqvJpPJuQsWLGiOx70fic6bw0kkEixYsKDZtu1vZVqjDEBVbWJMZQNR3sqX5K6jqs2WZU1avnx5y6FDpTDmlx0LnoenNvNeYwtzsj1n+fLlLZZl/cG27aWF1K2QbN26VV977bW3Lcv6arF18ZKGhgZWrFjRalnWlapaUAegCp+bRzSe5OequraQstqyBpJvwytx+Hcv5bZl586dbN26tTEWi80uph65sgd4DlpiMCV9JMO27Uej0eiaZ5991vOZubw5nNWrV8ePHDnyQjKZ/FlH6VR1PwlmsZBo4QcEskdV/2Lb9vcXLVrUnEyW0JhfO7z5Llz7S1qaWpmoqlnN/Wzfvp1du3YdtCzrhhOnLk0OHjyYanQnFrrRLSWSySQLFy5stm37u6q6rdDyfvYs9p9380ZzjG8XWlY6bwErnUbyKlUtWkVsampiyZIlLfF4fIqqlk8v1KUV+DVEEzBbVY9Zf0ZVNRaLzdyyZUvjrl0nXPoyr3TqNb/169db1dXVx41jJpNJ6uvrW+Lx+OfaWyguHbV1mYTl9zxOLf2K8G77m5kdbSKRuKehoWFiXV1dTY8e3nzn1tzc/MF3MNvfIjTnkRNP+AOs2kEymeQ/VPUvbeMSiUTwqaeeOu4627ZtS7oNdUYHpaq6atUqKxQKlexY9e7du5O2bd+qqu2Nf+rq1autcDhcsjbkwqFDhzh8+PBfksnkj1NhqugdS7BOrsrv3IKt8Ph6ElGLSelvAO4EXZLFCyldoR6SSbhWVfekwizLCmUqz4Vkz549djKZfEBV16XCbPAX2v58sQ80Bkts1SWZ4lX1kIhMXbx48dPDhg3L+cGjqampU+13p1aLFpGvdBC9pjM9L3edpJlk8c1IgVjS1vMDiEhvoBbvlrKIA4/ifMT3ebL/uOwI8Ju2k/7utgpfgoxfPNW3t+Cle+5E4CNZyi8Wh4HH2+vYiMiVwEBPNfIGG6hT1Q/eTBORy4DBBZK3SVU3pMkaBFxRIFnp/E1Vf58m1w/8E3i+jUkU+HXK4bofsc8GyuVbNQunnnQ4+iEinwGGdEWQqv4027SdcjgGg8FgMORK2S1IZzAYDIbyxDgcg8FgMHhCpxyOiFxXKEWKhbGpPDA2lQfGpvKhGHZ19gnn7/HGG5vKA2NTeWBsKh9K3uEYDAaDwZATxuEYDAaDwRM663AeKogWxcXYVB4Ym8oDY1P54Lld5jscg8FgMHiCGVIzGAwGgydk5XBE5LMisk1EXhGR34tIt0Ir1hXE4dcickta2HsisiXt9wU3vLeIPC0ifxWRehEZnXZOSdgtIjeKyA5Xv2Ui0kdE/CJyv4i8LCKvisi/pKUfIiLrXJs2isgZaXFfcsN3icgvRCRYDJvS9JmUvvGbiLzo6pfKp6+74RERmS8iL7n5MSntnFEi8mc3bpWI9CuSLcPcra43i8gmETnXDf92Wj7d7i6TUi5lb1abevO6iMRFpG+Z16m5rg5bRKRORE4uxzrVtq07gQ1XiEhDmzw7yY3LmC8dXS8nVLXDH9AbeAcY4h7fDfz8ROcV6wecCawGmoFb3LChwM520i8CvuP+Hw7sw1mHrCTsBs4F3gC6u8f3Ag8Cc4A/4izA2hN4Gfi4m2Yj8Hn3/2eAepw12j6GsyBvb5zORh3wjSLm1RDgVaDJPa7CWSstmCHtPcBD7v/T3HwaAIRcm8a4cTcAfyyCLRHgb8Bl7vFEN08uAza7tlUAa4Hp5VD2MtgYBJ4Hri/zOvUpYC8wwD2eCSwutzpF5rauIxvuSuVLm+u0my8dXS8nnbMw6gvAirTjgTjb+EqxK0A7+v4MZxHMx9Iy4RrgJeBPODvXfg/wuzcxCvROO38NMLmU7MZtgN0Gqw64E/gfYFpamtuBnwL9cRb29KXFvQGMAP4NmJcW/klgW5HyKQK8AFzJUYeTaghWAduBHwOVbtwuYGTa+Y8BNwNjgB1p4SGcFX1P8dieScD6tGMBzgEeBr6eFj4beLJcyl4bG28Flrn/y7ZOAbcAj6UdV7ll5rlyqlNkbusytgvu/7XASmCrm2/j3PB286Wj6+Xyy2ZI7VQ4Zru0vUA3SnTVVFW9UVXntwkO4NzoCcA44NPAXKAXTiF6Ny3tXpyec8nYrapxdwhpL47+j7ajX0rv/XrsKtId2TSggKp3xIPuL32F8ZNwKz0wEudJ5i43riN7PwhXVQt4F6eR8JKPAm+LyK9EZBNORQ201Y+jepdF2UshIr2ArwGprePLuU69AFwkIqmV0a/B6ah8mDKqU+20dR3pcxD4L5ynzm8DS0XkRPmSV/uycTg+Mu/lXfq7lLmo6sOqOldVm1X1MHAfcBWZbRMc20rKblV9QlV74fQwnuF4/TrSu724VLiniMgcIKGqj6SHq+qTqjpTVRvU2aL8Tpx8gtzs9ZIgzvDZQ6p6HvAAzlBEmK7nU4pi1rnrcJ5uXoPyrlOq+ifgDpwGdxPO1g8NlHGdSqNdfVR1sqouVof1wAZgfIZzUuTdvmwczh4cz5+iP3BIs9xlshQQkZkick56EM4+NO840XJyWtyHcbx4SdgtIoNFZGxa0CM4e9bsa6Nfut79UhPTGeIyneM1s4GRIrIFp1GudCcwrxaRcWnpUvkE7et+TLg7YXsKzv3xkv3AS6r6AoCqLsMZYrLJrHfJl7021OI8WQNlX6dOAtaq6gi3c7DMjeqojJV6nUqRUR8R6SEi32ljQyrPOsqX/NqXxThhH+AARyeU7gIe9WKMsovjm49xdFzzbuAJnAagEmdM+Vo37g/At9z/57i2di8Vu4FPuBncyz2ehTMGOxdYjjO00QNnPP1CN80mYIb7/9PAbpzOxTnA665tAvwOuK3I+TSQo3M4N+BMzla6eTUf+IEb9yPgF+7/ATgT9KfjPFnsBUa7cdcDzxXBjg/h9JLPdY/H4TS+V7j5UYXztPMccHU5lL0023riTEwH08LKuU4NdctMN/f45zjzIWVZpzi2rctog5tP+4Apbroat3z26ihfOronOemapUGX4TRyLwFPAScXq/DnmAkRnCeDv+JMPt/J0Y9e+7o3tB5novrSUrMbpyGuB1JPBIPcAnA/sMO16Za09EPcBqDerSgj0uKuccNfAX4DVBQ5nwZy1OH4gB+693sXzpfQYTeuGnjctfdl4Itp1/g4jqPaAawDBhbJlnE48wP1wIvAWDf8O2n5dG85lT1Xl5HAq23Cyr1O3ejq8ArwKxynWZZ1imPbuo5sOA/nLcNUW/KpE+VLR9fL5WdWGjAYDAaDJ5iVBgwGg8HgCcbhGAwGg8ETjMMxGAwGgycYh2MwGAwGTzAOx2AwGAyeYByOwWAwGDzBOByDwWAweIJxOAaDwWDwhP8DBF5zrbLRG2QAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 504x28.8 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mibig_domains = pd.read_csv('../data/training/positive/mibig_bgcs_all.csv').groupby('contig_id')\n",
    "for contig_id, sim in first_similarity_mibig[:5].iteritems():\n",
    "    bgc = mibig_domains.get_group(contig_id)\n",
    "    print('{}: Similarity {:.2f}%:'.format(contig_id, sim*100))\n",
    "    fig = plot_domain_genes(bgc, figure_width=7)\n",
    "    fig.set_figheight(0.4)\n",
    "    fig.axes[0].legend([]).remove()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 722,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Alkaloid</th>\n",
       "      <th>NRP</th>\n",
       "      <th>Other</th>\n",
       "      <th>Polyketide</th>\n",
       "      <th>RiPP</th>\n",
       "      <th>Saccharide</th>\n",
       "      <th>Terpene</th>\n",
       "      <th>classes</th>\n",
       "      <th>Antibacterial</th>\n",
       "      <th>Cytotoxic</th>\n",
       "      <th>...</th>\n",
       "      <th>candidate_id</th>\n",
       "      <th>avg_prediction</th>\n",
       "      <th>nucl_length</th>\n",
       "      <th>nucl_start</th>\n",
       "      <th>nucl_end</th>\n",
       "      <th>num_bio_domains</th>\n",
       "      <th>num_all_domains</th>\n",
       "      <th>num_proteins</th>\n",
       "      <th>protein_ids</th>\n",
       "      <th>pfam_ids</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>candidate_hash</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ffd47679ea075ec67ed6bb995467cbcb</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.268595</td>\n",
       "      <td>0.371381</td>\n",
       "      <td>0.340024</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.031667</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.694833</td>\n",
       "      <td>0.07</td>\n",
       "      <td>...</td>\n",
       "      <td>CP004886.1(314916-333380)</td>\n",
       "      <td>0.961393</td>\n",
       "      <td>18465</td>\n",
       "      <td>314916</td>\n",
       "      <td>333380</td>\n",
       "      <td>6</td>\n",
       "      <td>27</td>\n",
       "      <td>18</td>\n",
       "      <td>CP004886_368;CP004886_369;CP004886_370;CP00488...</td>\n",
       "      <td>PF13527;PF13508;PF13673;PF00583;PF02626;PF0262...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>307e7e7ef2b7256aa60260c3f20ed974</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.253262</td>\n",
       "      <td>0.287048</td>\n",
       "      <td>0.449690</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.021667</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.673167</td>\n",
       "      <td>0.04</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003234.1(311976-333375)</td>\n",
       "      <td>0.969206</td>\n",
       "      <td>21400</td>\n",
       "      <td>311976</td>\n",
       "      <td>333375</td>\n",
       "      <td>6</td>\n",
       "      <td>27</td>\n",
       "      <td>19</td>\n",
       "      <td>CP003234_326;CP003234_327;CP003234_330;CP00323...</td>\n",
       "      <td>PF02602;PF07690;PF13347;PF02626;PF02682;PF0149...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2b6c53e3e4812e6d8e813b9ab00cfb44</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.577595</td>\n",
       "      <td>0.162452</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.677333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003324.1(5117291-5135252)</td>\n",
       "      <td>0.960224</td>\n",
       "      <td>17962</td>\n",
       "      <td>5117291</td>\n",
       "      <td>5135252</td>\n",
       "      <td>7</td>\n",
       "      <td>17</td>\n",
       "      <td>12</td>\n",
       "      <td>CP003324_4800;CP003324_4801;CP003324_4803;CP00...</td>\n",
       "      <td>PF09995;PF00440;PF00668;PF07760;PF00903;PF1366...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f5e0ed012e76c5eda3c862febdee4e83</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.560095</td>\n",
       "      <td>0.169952</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.687333</td>\n",
       "      <td>0.04</td>\n",
       "      <td>...</td>\n",
       "      <td>AP012555.1(4575446-4595886)</td>\n",
       "      <td>0.964836</td>\n",
       "      <td>20441</td>\n",
       "      <td>4575446</td>\n",
       "      <td>4595886</td>\n",
       "      <td>7</td>\n",
       "      <td>21</td>\n",
       "      <td>15</td>\n",
       "      <td>AP012555_4218;AP012555_4219;AP012555_4220;AP01...</td>\n",
       "      <td>PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>257ddd67560ffe89405860edcfeec6b5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.567595</td>\n",
       "      <td>0.162452</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.677333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP000479.1(5011141-5031539)</td>\n",
       "      <td>0.962017</td>\n",
       "      <td>20399</td>\n",
       "      <td>5011141</td>\n",
       "      <td>5031539</td>\n",
       "      <td>7</td>\n",
       "      <td>20</td>\n",
       "      <td>14</td>\n",
       "      <td>CP000479_4775;CP000479_4776;CP000479_4777;CP00...</td>\n",
       "      <td>PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>257ddd67560ffe89405860edcfeec6b5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.567595</td>\n",
       "      <td>0.162452</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.677333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP009482.1(4797621-4817593)</td>\n",
       "      <td>0.961966</td>\n",
       "      <td>19973</td>\n",
       "      <td>4797621</td>\n",
       "      <td>4817593</td>\n",
       "      <td>7</td>\n",
       "      <td>20</td>\n",
       "      <td>14</td>\n",
       "      <td>CP009482_4519;CP009482_4520;CP009482_4521;CP00...</td>\n",
       "      <td>PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>257ddd67560ffe89405860edcfeec6b5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.567595</td>\n",
       "      <td>0.162452</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.677333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP009493.1(4782355-4802330)</td>\n",
       "      <td>0.964243</td>\n",
       "      <td>19976</td>\n",
       "      <td>4782355</td>\n",
       "      <td>4802330</td>\n",
       "      <td>7</td>\n",
       "      <td>20</td>\n",
       "      <td>14</td>\n",
       "      <td>CP009493_4522;CP009493_4523;CP009493_4524;CP00...</td>\n",
       "      <td>PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>257ddd67560ffe89405860edcfeec6b5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.567595</td>\n",
       "      <td>0.162452</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.677333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP009614.1(4834304-4854702)</td>\n",
       "      <td>0.962074</td>\n",
       "      <td>20399</td>\n",
       "      <td>4834304</td>\n",
       "      <td>4854702</td>\n",
       "      <td>7</td>\n",
       "      <td>20</td>\n",
       "      <td>14</td>\n",
       "      <td>CP009614_4558;CP009614_4559;CP009614_4560;CP00...</td>\n",
       "      <td>PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62e36433ceeb2caaaeab262312897ffb</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.210262</td>\n",
       "      <td>0.270298</td>\n",
       "      <td>0.519440</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.021667</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>0.650000</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>AE000516.2(323452-331781)</td>\n",
       "      <td>0.944847</td>\n",
       "      <td>8330</td>\n",
       "      <td>323452</td>\n",
       "      <td>331781</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>AE000516_288;AE000516_289;AE000516_290;AE00051...</td>\n",
       "      <td>PF01896;PF13193;PF00501;PF02771;PF02770;PF0802...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>205ed77242499c881cc0c1c3cdf71f89</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.577595</td>\n",
       "      <td>0.162452</td>\n",
       "      <td>0.259952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.687333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP002275.1(5200177-5216363)</td>\n",
       "      <td>0.955887</td>\n",
       "      <td>16187</td>\n",
       "      <td>5200177</td>\n",
       "      <td>5216363</td>\n",
       "      <td>7</td>\n",
       "      <td>15</td>\n",
       "      <td>10</td>\n",
       "      <td>CP002275_4819;CP002275_4821;CP002275_4822;CP00...</td>\n",
       "      <td>PF00440;PF00668;PF07760;PF00903;PF13669;PF0044...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6ba88b1a3182d0183371fb410f9ce219</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.567595</td>\n",
       "      <td>0.152452</td>\n",
       "      <td>0.279952</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.687333</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003491.1(5217503-5232503)</td>\n",
       "      <td>0.956674</td>\n",
       "      <td>15001</td>\n",
       "      <td>5217503</td>\n",
       "      <td>5232503</td>\n",
       "      <td>7</td>\n",
       "      <td>14</td>\n",
       "      <td>9</td>\n",
       "      <td>CP003491_4814;CP003491_4816;CP003491_4817;CP00...</td>\n",
       "      <td>PF00440;PF00668;PF07760;PF00903;PF13669;PF0044...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6e6077b2b45de59b3bc8174e15ed36e1</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.198452</td>\n",
       "      <td>0.267845</td>\n",
       "      <td>0.523702</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.011667</td>\n",
       "      <td>0.01</td>\n",
       "      <td>Polyketide</td>\n",
       "      <td>0.670000</td>\n",
       "      <td>0.03</td>\n",
       "      <td>...</td>\n",
       "      <td>CP005387.1(324510-330261)</td>\n",
       "      <td>0.936981</td>\n",
       "      <td>5752</td>\n",
       "      <td>324510</td>\n",
       "      <td>330261</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>5</td>\n",
       "      <td>CP005387_297;CP005387_298;CP005387_299;CP00538...</td>\n",
       "      <td>PF13193;PF00501;PF00441;PF02770;PF02771;PF0277...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad6a07c13956c21f4ea44f40fc801d58</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.558429</td>\n",
       "      <td>0.173119</td>\n",
       "      <td>0.268452</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003323.1(5024974-5036557)</td>\n",
       "      <td>0.948203</td>\n",
       "      <td>11584</td>\n",
       "      <td>5024974</td>\n",
       "      <td>5036557</td>\n",
       "      <td>7</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>CP003323_4701;CP003323_4702;CP003323_4703;CP00...</td>\n",
       "      <td>PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad6a07c13956c21f4ea44f40fc801d58</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.558429</td>\n",
       "      <td>0.173119</td>\n",
       "      <td>0.268452</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>CP003347.1(5066697-5078313)</td>\n",
       "      <td>0.955510</td>\n",
       "      <td>11617</td>\n",
       "      <td>5066697</td>\n",
       "      <td>5078313</td>\n",
       "      <td>7</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>CP003347_4701;CP003347_4702;CP003347_4703;CP00...</td>\n",
       "      <td>PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ad6a07c13956c21f4ea44f40fc801d58</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.558429</td>\n",
       "      <td>0.173119</td>\n",
       "      <td>0.268452</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.005000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NRP</td>\n",
       "      <td>0.700000</td>\n",
       "      <td>0.01</td>\n",
       "      <td>...</td>\n",
       "      <td>CP009499.1(4800724-4812303)</td>\n",
       "      <td>0.955701</td>\n",
       "      <td>11580</td>\n",
       "      <td>4800724</td>\n",
       "      <td>4812303</td>\n",
       "      <td>7</td>\n",
       "      <td>12</td>\n",
       "      <td>7</td>\n",
       "      <td>CP009499_4478;CP009499_4479;CP009499_4480;CP00...</td>\n",
       "      <td>PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  Alkaloid       NRP     Other  Polyketide  \\\n",
       "candidate_hash                                                               \n",
       "ffd47679ea075ec67ed6bb995467cbcb      0.01  0.268595  0.371381    0.340024   \n",
       "307e7e7ef2b7256aa60260c3f20ed974      0.00  0.253262  0.287048    0.449690   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44      0.00  0.577595  0.162452    0.259952   \n",
       "f5e0ed012e76c5eda3c862febdee4e83      0.00  0.560095  0.169952    0.259952   \n",
       "257ddd67560ffe89405860edcfeec6b5      0.00  0.567595  0.162452    0.259952   \n",
       "257ddd67560ffe89405860edcfeec6b5      0.00  0.567595  0.162452    0.259952   \n",
       "257ddd67560ffe89405860edcfeec6b5      0.00  0.567595  0.162452    0.259952   \n",
       "257ddd67560ffe89405860edcfeec6b5      0.00  0.567595  0.162452    0.259952   \n",
       "62e36433ceeb2caaaeab262312897ffb      0.00  0.210262  0.270298    0.519440   \n",
       "205ed77242499c881cc0c1c3cdf71f89      0.00  0.577595  0.162452    0.259952   \n",
       "6ba88b1a3182d0183371fb410f9ce219      0.00  0.567595  0.152452    0.279952   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1      0.00  0.198452  0.267845    0.523702   \n",
       "ad6a07c13956c21f4ea44f40fc801d58      0.00  0.558429  0.173119    0.268452   \n",
       "ad6a07c13956c21f4ea44f40fc801d58      0.00  0.558429  0.173119    0.268452   \n",
       "ad6a07c13956c21f4ea44f40fc801d58      0.00  0.558429  0.173119    0.268452   \n",
       "\n",
       "                                  RiPP  Saccharide  Terpene     classes  \\\n",
       "candidate_hash                                                            \n",
       "ffd47679ea075ec67ed6bb995467cbcb  0.00    0.031667     0.01         NaN   \n",
       "307e7e7ef2b7256aa60260c3f20ed974  0.01    0.021667     0.00         NaN   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44  0.00    0.005000     0.00         NRP   \n",
       "f5e0ed012e76c5eda3c862febdee4e83  0.00    0.005000     0.01         NRP   \n",
       "257ddd67560ffe89405860edcfeec6b5  0.00    0.005000     0.01         NRP   \n",
       "257ddd67560ffe89405860edcfeec6b5  0.00    0.005000     0.01         NRP   \n",
       "257ddd67560ffe89405860edcfeec6b5  0.00    0.005000     0.01         NRP   \n",
       "257ddd67560ffe89405860edcfeec6b5  0.00    0.005000     0.01         NRP   \n",
       "62e36433ceeb2caaaeab262312897ffb  0.00    0.021667     0.00  Polyketide   \n",
       "205ed77242499c881cc0c1c3cdf71f89  0.00    0.005000     0.00         NRP   \n",
       "6ba88b1a3182d0183371fb410f9ce219  0.00    0.005000     0.00         NRP   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1  0.00    0.011667     0.01  Polyketide   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  0.00    0.005000     0.00         NRP   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  0.00    0.005000     0.00         NRP   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  0.00    0.005000     0.00         NRP   \n",
       "\n",
       "                                  Antibacterial  Cytotoxic  \\\n",
       "candidate_hash                                               \n",
       "ffd47679ea075ec67ed6bb995467cbcb       0.694833       0.07   \n",
       "307e7e7ef2b7256aa60260c3f20ed974       0.673167       0.04   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44       0.677333       0.03   \n",
       "f5e0ed012e76c5eda3c862febdee4e83       0.687333       0.04   \n",
       "257ddd67560ffe89405860edcfeec6b5       0.677333       0.03   \n",
       "257ddd67560ffe89405860edcfeec6b5       0.677333       0.03   \n",
       "257ddd67560ffe89405860edcfeec6b5       0.677333       0.03   \n",
       "257ddd67560ffe89405860edcfeec6b5       0.677333       0.03   \n",
       "62e36433ceeb2caaaeab262312897ffb       0.650000       0.03   \n",
       "205ed77242499c881cc0c1c3cdf71f89       0.687333       0.03   \n",
       "6ba88b1a3182d0183371fb410f9ce219       0.687333       0.03   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1       0.670000       0.03   \n",
       "ad6a07c13956c21f4ea44f40fc801d58       0.700000       0.01   \n",
       "ad6a07c13956c21f4ea44f40fc801d58       0.700000       0.01   \n",
       "ad6a07c13956c21f4ea44f40fc801d58       0.700000       0.01   \n",
       "\n",
       "                                                        ...                          \\\n",
       "candidate_hash                                          ...                           \n",
       "ffd47679ea075ec67ed6bb995467cbcb                        ...                           \n",
       "307e7e7ef2b7256aa60260c3f20ed974                        ...                           \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44                        ...                           \n",
       "f5e0ed012e76c5eda3c862febdee4e83                        ...                           \n",
       "257ddd67560ffe89405860edcfeec6b5                        ...                           \n",
       "257ddd67560ffe89405860edcfeec6b5                        ...                           \n",
       "257ddd67560ffe89405860edcfeec6b5                        ...                           \n",
       "257ddd67560ffe89405860edcfeec6b5                        ...                           \n",
       "62e36433ceeb2caaaeab262312897ffb                        ...                           \n",
       "205ed77242499c881cc0c1c3cdf71f89                        ...                           \n",
       "6ba88b1a3182d0183371fb410f9ce219                        ...                           \n",
       "6e6077b2b45de59b3bc8174e15ed36e1                        ...                           \n",
       "ad6a07c13956c21f4ea44f40fc801d58                        ...                           \n",
       "ad6a07c13956c21f4ea44f40fc801d58                        ...                           \n",
       "ad6a07c13956c21f4ea44f40fc801d58                        ...                           \n",
       "\n",
       "                                                 candidate_id  avg_prediction  \\\n",
       "candidate_hash                                                                  \n",
       "ffd47679ea075ec67ed6bb995467cbcb    CP004886.1(314916-333380)        0.961393   \n",
       "307e7e7ef2b7256aa60260c3f20ed974    CP003234.1(311976-333375)        0.969206   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44  CP003324.1(5117291-5135252)        0.960224   \n",
       "f5e0ed012e76c5eda3c862febdee4e83  AP012555.1(4575446-4595886)        0.964836   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP000479.1(5011141-5031539)        0.962017   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP009482.1(4797621-4817593)        0.961966   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP009493.1(4782355-4802330)        0.964243   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP009614.1(4834304-4854702)        0.962074   \n",
       "62e36433ceeb2caaaeab262312897ffb    AE000516.2(323452-331781)        0.944847   \n",
       "205ed77242499c881cc0c1c3cdf71f89  CP002275.1(5200177-5216363)        0.955887   \n",
       "6ba88b1a3182d0183371fb410f9ce219  CP003491.1(5217503-5232503)        0.956674   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1    CP005387.1(324510-330261)        0.936981   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  CP003323.1(5024974-5036557)        0.948203   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  CP003347.1(5066697-5078313)        0.955510   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  CP009499.1(4800724-4812303)        0.955701   \n",
       "\n",
       "                                 nucl_length nucl_start  nucl_end  \\\n",
       "candidate_hash                                                      \n",
       "ffd47679ea075ec67ed6bb995467cbcb       18465     314916    333380   \n",
       "307e7e7ef2b7256aa60260c3f20ed974       21400     311976    333375   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44       17962    5117291   5135252   \n",
       "f5e0ed012e76c5eda3c862febdee4e83       20441    4575446   4595886   \n",
       "257ddd67560ffe89405860edcfeec6b5       20399    5011141   5031539   \n",
       "257ddd67560ffe89405860edcfeec6b5       19973    4797621   4817593   \n",
       "257ddd67560ffe89405860edcfeec6b5       19976    4782355   4802330   \n",
       "257ddd67560ffe89405860edcfeec6b5       20399    4834304   4854702   \n",
       "62e36433ceeb2caaaeab262312897ffb        8330     323452    331781   \n",
       "205ed77242499c881cc0c1c3cdf71f89       16187    5200177   5216363   \n",
       "6ba88b1a3182d0183371fb410f9ce219       15001    5217503   5232503   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1        5752     324510    330261   \n",
       "ad6a07c13956c21f4ea44f40fc801d58       11584    5024974   5036557   \n",
       "ad6a07c13956c21f4ea44f40fc801d58       11617    5066697   5078313   \n",
       "ad6a07c13956c21f4ea44f40fc801d58       11580    4800724   4812303   \n",
       "\n",
       "                                 num_bio_domains num_all_domains  \\\n",
       "candidate_hash                                                     \n",
       "ffd47679ea075ec67ed6bb995467cbcb               6              27   \n",
       "307e7e7ef2b7256aa60260c3f20ed974               6              27   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44               7              17   \n",
       "f5e0ed012e76c5eda3c862febdee4e83               7              21   \n",
       "257ddd67560ffe89405860edcfeec6b5               7              20   \n",
       "257ddd67560ffe89405860edcfeec6b5               7              20   \n",
       "257ddd67560ffe89405860edcfeec6b5               7              20   \n",
       "257ddd67560ffe89405860edcfeec6b5               7              20   \n",
       "62e36433ceeb2caaaeab262312897ffb               6              12   \n",
       "205ed77242499c881cc0c1c3cdf71f89               7              15   \n",
       "6ba88b1a3182d0183371fb410f9ce219               7              14   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1               4               8   \n",
       "ad6a07c13956c21f4ea44f40fc801d58               7              12   \n",
       "ad6a07c13956c21f4ea44f40fc801d58               7              12   \n",
       "ad6a07c13956c21f4ea44f40fc801d58               7              12   \n",
       "\n",
       "                                  num_proteins  \\\n",
       "candidate_hash                                   \n",
       "ffd47679ea075ec67ed6bb995467cbcb            18   \n",
       "307e7e7ef2b7256aa60260c3f20ed974            19   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44            12   \n",
       "f5e0ed012e76c5eda3c862febdee4e83            15   \n",
       "257ddd67560ffe89405860edcfeec6b5            14   \n",
       "257ddd67560ffe89405860edcfeec6b5            14   \n",
       "257ddd67560ffe89405860edcfeec6b5            14   \n",
       "257ddd67560ffe89405860edcfeec6b5            14   \n",
       "62e36433ceeb2caaaeab262312897ffb             7   \n",
       "205ed77242499c881cc0c1c3cdf71f89            10   \n",
       "6ba88b1a3182d0183371fb410f9ce219             9   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1             5   \n",
       "ad6a07c13956c21f4ea44f40fc801d58             7   \n",
       "ad6a07c13956c21f4ea44f40fc801d58             7   \n",
       "ad6a07c13956c21f4ea44f40fc801d58             7   \n",
       "\n",
       "                                                                        protein_ids  \\\n",
       "candidate_hash                                                                        \n",
       "ffd47679ea075ec67ed6bb995467cbcb  CP004886_368;CP004886_369;CP004886_370;CP00488...   \n",
       "307e7e7ef2b7256aa60260c3f20ed974  CP003234_326;CP003234_327;CP003234_330;CP00323...   \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44  CP003324_4800;CP003324_4801;CP003324_4803;CP00...   \n",
       "f5e0ed012e76c5eda3c862febdee4e83  AP012555_4218;AP012555_4219;AP012555_4220;AP01...   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP000479_4775;CP000479_4776;CP000479_4777;CP00...   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP009482_4519;CP009482_4520;CP009482_4521;CP00...   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP009493_4522;CP009493_4523;CP009493_4524;CP00...   \n",
       "257ddd67560ffe89405860edcfeec6b5  CP009614_4558;CP009614_4559;CP009614_4560;CP00...   \n",
       "62e36433ceeb2caaaeab262312897ffb  AE000516_288;AE000516_289;AE000516_290;AE00051...   \n",
       "205ed77242499c881cc0c1c3cdf71f89  CP002275_4819;CP002275_4821;CP002275_4822;CP00...   \n",
       "6ba88b1a3182d0183371fb410f9ce219  CP003491_4814;CP003491_4816;CP003491_4817;CP00...   \n",
       "6e6077b2b45de59b3bc8174e15ed36e1  CP005387_297;CP005387_298;CP005387_299;CP00538...   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  CP003323_4701;CP003323_4702;CP003323_4703;CP00...   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  CP003347_4701;CP003347_4702;CP003347_4703;CP00...   \n",
       "ad6a07c13956c21f4ea44f40fc801d58  CP009499_4478;CP009499_4479;CP009499_4480;CP00...   \n",
       "\n",
       "                                                                           pfam_ids  \n",
       "candidate_hash                                                                       \n",
       "ffd47679ea075ec67ed6bb995467cbcb  PF13527;PF13508;PF13673;PF00583;PF02626;PF0262...  \n",
       "307e7e7ef2b7256aa60260c3f20ed974  PF02602;PF07690;PF13347;PF02626;PF02682;PF0149...  \n",
       "2b6c53e3e4812e6d8e813b9ab00cfb44  PF09995;PF00440;PF00668;PF07760;PF00903;PF1366...  \n",
       "f5e0ed012e76c5eda3c862febdee4e83  PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...  \n",
       "257ddd67560ffe89405860edcfeec6b5  PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...  \n",
       "257ddd67560ffe89405860edcfeec6b5  PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...  \n",
       "257ddd67560ffe89405860edcfeec6b5  PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...  \n",
       "257ddd67560ffe89405860edcfeec6b5  PF00823;PF09995;PF00440;PF00668;PF14304;PF0090...  \n",
       "62e36433ceeb2caaaeab262312897ffb  PF01896;PF13193;PF00501;PF02771;PF02770;PF0802...  \n",
       "205ed77242499c881cc0c1c3cdf71f89  PF00440;PF00668;PF07760;PF00903;PF13669;PF0044...  \n",
       "6ba88b1a3182d0183371fb410f9ce219  PF00440;PF00668;PF07760;PF00903;PF13669;PF0044...  \n",
       "6e6077b2b45de59b3bc8174e15ed36e1  PF13193;PF00501;PF00441;PF02770;PF02771;PF0277...  \n",
       "ad6a07c13956c21f4ea44f40fc801d58  PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...  \n",
       "ad6a07c13956c21f4ea44f40fc801d58  PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...  \n",
       "ad6a07c13956c21f4ea44f40fc801d58  PF00668;PF07760;PF00903;PF13669;PF00440;PF0277...  \n",
       "\n",
       "[15 rows x 31 columns]"
      ]
     },
     "execution_count": 722,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "deepbgc.set_index('candidate_hash').loc[first_similarity[:10].index]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Overlap Venn diagram"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 723,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_single_contig_overlap(a_cands, b_cands):\n",
    "    overlaps = []\n",
    "    for c, cand in a_cands.iterrows():\n",
    "        # For each suitable candidate from other model\n",
    "        overlap = False\n",
    "        for i, other in b_cands.iterrows():\n",
    "            if other['nucl_start'] < cand['nucl_end'] and other['nucl_end'] > cand['nucl_start']:\n",
    "                overlap = True\n",
    "                break\n",
    "        overlaps.append(overlap)\n",
    "    return pd.Series(overlaps, index=a_cands.index)\n",
    "\n",
    "def get_overlap(a_cands, b_cands):\n",
    "    a_grouped = a_cands.groupby('contig_id')\n",
    "    b_grouped = b_cands.groupby('contig_id')\n",
    "\n",
    "    overlaps = []\n",
    "    # Get overlap separately for all contigs\n",
    "    for contig_id in a_grouped.groups:\n",
    "        a_contig_cands = a_grouped.get_group(contig_id)\n",
    "        b_contig_cands = b_grouped.get_group(contig_id) if contig_id in b_grouped.groups else pd.DataFrame()\n",
    "        overlaps.append(get_single_contig_overlap(a_contig_cands, b_contig_cands))\n",
    "    return pd.concat(overlaps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 724,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_multiple_overlaps(candidates):\n",
    "    overlaps = []\n",
    "    for a_model, a_cands in candidates.groupby('model'):\n",
    "        overlap = pd.DataFrame()\n",
    "        for b_model, b_cands in candidates.groupby('model'):\n",
    "            if a_model == b_model:\n",
    "                continue\n",
    "            overlap[b_model] = get_overlap(a_cands, b_cands)        \n",
    "        overlap['model'] = a_model\n",
    "        overlap[a_model] = True\n",
    "        overlaps.append(overlap)\n",
    "    return pd.concat(overlaps)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 725,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ClusterFinder</th>\n",
       "      <th>DeepBGC</th>\n",
       "      <th>antiSMASH</th>\n",
       "      <th>model</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>ClusterFinder</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>ClusterFinder</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>ClusterFinder</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>ClusterFinder</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>ClusterFinder</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   ClusterFinder  DeepBGC  antiSMASH          model\n",
       "0           True    False      False  ClusterFinder\n",
       "1           True    False      False  ClusterFinder\n",
       "2           True    False      False  ClusterFinder\n",
       "3           True    False      False  ClusterFinder\n",
       "4           True    False      False  ClusterFinder"
      ]
     },
     "execution_count": 725,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "overlaps = get_multiple_overlaps(candidates)\n",
    "overlaps.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 726,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "566"
      ]
     },
     "execution_count": 726,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x = deepbgc[deepbgc['coverage'] == 0].set_index('candidate_id')\n",
    "len(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 727,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "model\n",
       "ClusterFinder    34\n",
       "DeepBGC           9\n",
       "antiSMASH         3\n",
       "dtype: int64"
      ]
     },
     "execution_count": 727,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "candidates[candidates['contig_id'] == 'CP002955.1'].groupby('model').size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 728,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CP002955.1 of length 1919915\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x54 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plot_candidate_region('002d6748dad4145b1f09ca9dc93ead44', start=500000, end=2000000)\n",
    "fig.set_figwidth(7)\n",
    "fig.set_figheight(0.75)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 729,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>model</th>\n",
       "      <th>All overlap</th>\n",
       "      <th>Novel</th>\n",
       "      <th>ClusterFinder overlap</th>\n",
       "      <th>DeepBGC overlap</th>\n",
       "      <th>antiSMASH overlap</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ClusterFinder</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           model  All overlap  Novel  ClusterFinder overlap  DeepBGC overlap  \\\n",
       "0  ClusterFinder        False   True                  False            False   \n",
       "1  ClusterFinder        False   True                  False            False   \n",
       "2  ClusterFinder        False   True                  False            False   \n",
       "3  ClusterFinder        False   True                  False            False   \n",
       "4  ClusterFinder        False   True                  False            False   \n",
       "\n",
       "   antiSMASH overlap  \n",
       "0              False  \n",
       "1              False  \n",
       "2              False  \n",
       "3              False  \n",
       "4              False  "
      ]
     },
     "execution_count": 729,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "overlap_bins = overlaps[['model']].copy()\n",
    "num_overlap = overlaps.sum(axis=1)\n",
    "overlap_bins['All overlap'] = num_overlap == 3\n",
    "overlap_bins['Novel'] = num_overlap == 1\n",
    "novel_or_all = overlap_bins['All overlap'] | overlap_bins['Novel']\n",
    "overlap_bins['ClusterFinder overlap'] = overlaps['ClusterFinder'] & ~novel_or_all & (overlaps['model'] != 'ClusterFinder')\n",
    "overlap_bins['DeepBGC overlap'] = overlaps['DeepBGC'] & ~novel_or_all & (overlaps['model'] != 'DeepBGC')\n",
    "overlap_bins['antiSMASH overlap'] = overlaps['antiSMASH'] & ~novel_or_all & (overlaps['model'] != 'antiSMASH')\n",
    "overlap_bins.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 730,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>All overlap</th>\n",
       "      <th>Novel</th>\n",
       "      <th>ClusterFinder overlap</th>\n",
       "      <th>DeepBGC overlap</th>\n",
       "      <th>antiSMASH overlap</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ClusterFinder</th>\n",
       "      <td>4241</td>\n",
       "      <td>47074</td>\n",
       "      <td>0</td>\n",
       "      <td>6114</td>\n",
       "      <td>5062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DeepBGC</th>\n",
       "      <td>3862</td>\n",
       "      <td>566</td>\n",
       "      <td>6062</td>\n",
       "      <td>0</td>\n",
       "      <td>436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>antiSMASH</th>\n",
       "      <td>3646</td>\n",
       "      <td>6158</td>\n",
       "      <td>3911</td>\n",
       "      <td>150</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               All overlap  Novel  ClusterFinder overlap  DeepBGC overlap  \\\n",
       "model                                                                       \n",
       "ClusterFinder         4241  47074                      0             6114   \n",
       "DeepBGC               3862    566                   6062                0   \n",
       "antiSMASH             3646   6158                   3911              150   \n",
       "\n",
       "               antiSMASH overlap  \n",
       "model                             \n",
       "ClusterFinder               5062  \n",
       "DeepBGC                      436  \n",
       "antiSMASH                      0  "
      ]
     },
     "execution_count": 730,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "overlap_counts = overlap_bins.groupby('model').sum().astype(np.int)\n",
    "overlap_counts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 733,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "bin_colors = {\n",
    "    'All overlap': '#444444',\n",
    "    'Novel': '#00dd44',\n",
    "    'ClusterFinder overlap': 'dodgerblue',\n",
    "    'antiSMASH overlap': 'orange',\n",
    "    'DeepBGC overlap': 'red',\n",
    "    'ClusterFinder&DeepBGC overlap': 'purple'\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 747,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAJMAAACCCAYAAABCb/orAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAEshJREFUeJztnXuUFNWdxz+/GYbHDAgIaEChh5fiE5REMOATyihBY1wwUdfHkrg+YxtQo2dXk6Mmq4loWnfXuFF0faEbUUGjxEJEUVAE5akgBtOAvAQEBpB5/vaPW4NNM4/umbpV3Ux9zukz09VV996a/s69Vb/6PURViYjwg4KwBxBx4BCJKcI3IjFF+EYkpgjfiMQU4RuRmCJ8IxJThG9EYorwjUhMEb4RiSnCNyIxRfhGJKYI34jEFOEbkZgifCMSU4RvRGKK8I1ITBG+kZGYRGSiiKwWkYXe63lv+wIR+SRl+831HD9FRD5P2e8Bb3tf7/1KERmXsv+lInKXHycYERySiduuiMwFJqjqnJRtJcCXQDdVrWzk+HXAd1V1Xdr2+4EPgReBT1S1r4h0AFzgTFXdne0JRYRHozOTiLQBTgBuEZEl3izTCzgJ2AlM97Y/ICLt6ji+N9AB+LO33+MicrD3cTnQHigBarxtvwHui4SUfzQ6M3li+C/gFmAZcBNwMeZLHwPEgd3AM8AaVb0x7fghwK3ADcA64I9AT1U9X0S6A08B3YBfAyuAB1T1bJ/Oby+xBG2Ao4BSr78uGCEXA5WYf4ydQBnwNfA5sCIZZ4/fYzlQyWiZ2+cAEQG2AwNV9YuU7YOBF1U11sjxXYANQImqVqR99jowAegNXAvsAK5T1a3ZjDGWoBAYCowAjgeOBfoBhdm0g5ktk8ByYAFm+Z2bjNPgst5SyWRmOh4jnKe894L5kq8Fkqr6jrf9u8Bzqtov7fhTgM6qOs173xVzrVWsqtUp+40FTlbV8SKyArO0XgAcqaq3N3YisQQHA6OBUcBZQOcMzr8p7ARmAX8DXkjG2WCpn7yjVQb71AAPisi73kx0DbAYs0TcJyKnARXAeOD5Oo5vDzzkHb8VuBl4IU1Ixd72kd6mIq/fGswyVC+xBCd7YxoLtM3gfJpLe4xoRwN/jCV4A/gzMC0Zp7rBIw9wMr2b+2fMdU8hsBb4mffzXswftRXwFvALVS0XkfOAq1V1lHf8BODnmAv+JcCVqvp1Svt3AytV9X+999dilrttwJjU5RQglqA1cDlwHTCwyWfvL2uB+4GHW+p1VtbXTGESS1AAXIa5+G/w2ixE1gJ3AZOScarCHkyQ5I2YYglOx9wJ5spM1BgrgVuScV4OeyBBkfNiiiUoASYCV4U9liYyGbg+GSerO9J8JKfFFEswFGOH6tfYvjnOBuCqZJxpYQ/EJjkpplgCwRgx/53sbUO5zMPADQfqtVTOiSmWoC3wJOZW/0BkBjA2GWdb2APxm5wSUyxBF2AqMCzssVhmBTA6GefzsAfiJzkjpliC3hircv+wxxIQWwAnGefjsAfiFzkhpliC7sB7mGdyLYktwJnJOIvDHogfhO5pGUvQGTMjtTQhgfFccGOJA2M2DlVMsQTFwKvAcWGOI2QOAd6IJfhO2ANpLmHPTE8D3w95DLlAKfC85zqTt4QmpliCm4Efh9V/DnIq8NuwB9EcQrkA9yzbs8nMBaYlocCPknFeCXsgTSFwMcUSdAAWAn0C7Th/2AYck4yzrtE9c4wwlrm7iITUEJ0wflF5R6AzUyzB0cAiouUtE0Yk48wMexDZELSYXL51zQ2DPRj/7VkYJ7YNGJ/uckw41gBMBEvtz9IQxljLcuD4fApeCGyGiCX4MeEIaT0wBXgdeCsZ55sG9p2d+iaW4BhMWNcVQA9bA6yHAZjwsIkB99tkApmZPJeSZZj/9qBYD/wH8D/JOOXNaSiWoAgTI3g7wZ7DBqC0ueMPiqAuwM8huC+hArgH6J+M85AfX0QyTmUyzmRM+NXvIbAolO9gfN7zgqBmphmYgEjbbADOTcaZb7MTz072BHCkzX48VgIDkvG94fM5i/WZKZbgeIIR0ifAUNtCAkjGeR8YRN1xgn7THzg/gH6aTRDL3PUB9PEWMCwZJxlAXwB4sXGXYJ4v2ubKAPpoNlaXOe/B5Qagq7VOzB3YyGScikb3tIB3ji8C51nsphrokYyzyWIfzcb2zHQadoX0JcafOhQhAXgh4T/F5JnyFUW3V3RY9d7mQfd+tOac837kd/t+Y9vOdIHFthW4JBlno8U+MiIZ55tYgnHAxzTzb6poWWX75JKyvlNa7TrszYEUVtb6w2/G5DSoExGZiAnCqI3PW6GqP0nb5wcYz4RWmDwOt6rqGyLSGngZOAJ4U1Wv8vbvCzyiqhnZB60tc55taQ1wmJUO4IFknPGW2m4SsQQTIfsxKbqrqmTtorI+Uwp39nQHUlhRVwKOcqCr9nF31tVGXdn90j7vCKwCTlXVZV52m3eAnsAZwA9V9SoRmQ7cpKpLRWQqcIeqLsrkPGzOTAOwJ6QdmHwDucZvMEteo9ZyRXdXlaxbVNb7RdnVa/pALaxozEmwDeaueGr6B2nZ/foCnwG/VNXVKbsVAdeq6jLv/SeAYC5DyoESb4YqBipEZDQmeVtGQgK7YjrZYtuPJePssNh+k0jGKYslmIAJCd8PRfdUF29YVNb75eqdvV4bqK32ZPs3GkwdYsKIdyYmaLU2u99UETlRvaVHVTezrynjTuAzVf1CRJKYJXIh8BdMgrMnMcbmjLEpppMstVsNJCy17Qf/B9yHNysrWl7dbtOistKplTtjfz1ei3YPaUbbJ9a10Us5NKr2vYjch3n0Uwrsk45IRFphXFzOwbP/qWoNJuVR7T63A48BXUVkEmZWu11VGwzLsimmQZbanRKkPSlbknFqej1YOam6zbazd5a+UlFWOu1YLdrl1z/WCXVtTM/uV7sZ9vU4EJHOwAveZ0NVdUsdbfXCPJA/A2NDux8jyMnAKQ0NzqaYjrXU7qOW2vWNNaNHvYCZGfymh6xyOmgftyxte53Z/VR1be0OIlIIvIZZBq9W1fryHUwEfqWqNd61WBUZZPADS3YmLxauxELTlZhgzZxG+7iLwdrs2Xe//lSXAr8AXhGRTzGBGheJSA8vaX8P4EJM0tghwPyUBP97w8xEZCSwU1Xf9zZNBB7HPGFo9J/DimkgluAozN2C33yQjDPUQru+I6uc5zFfoN+M0T7uFAvtNhtbFnBbAYX5FEad8S11luRq+sW8E5ON2c4WtjKc7FcFIlewJaZOltrNp5xGtkKVWltqt9nYupuz5YmYN1Et7aqrv2mtNTYMqzkbQm7ry7GVZi9vxLT7w7cKgIMsNN2KIyy06gO2lrkWLyYysMs0kcbKsZ0vIul2qNrPhojIhyLyqYi86RVCQkS6iMh7Xk3A21P2P0VEnsx0YPkmppy9XqgDWw+56/XdEpH+mEc5UsdnrTHW7xtV9Sjv98e8jy/BpDbqD/xURA7yjJy/A36V6cBsiWk/M71PHG2pXRvYSjddp7elV3/maep3gfkesENVa42+jwEjvCpbtXX/2mJm/2pMoaWpqro+04HZEpMt6+9gS+3awNazyS/r2f6I96rPFtcT418GgFee7SvMDPo0JhRtPuZZXDFwEVk+ULclpn9gPCH95rhYwprZwTccx5HJi0vbbihrO0+V7T43vyZ9g1e4qEpVJzVwXAH7fycCVKvqLlW9QFWPUdVHMMGrvwZOFJFXReQlr4hlg1i5oE3G2RNLsBH/jZeFGLeJnHyckMLgSQv6nTNpQT8ErTnmkG3LR/Zbv3HI4VtKuhSXDxChfRPbVeo2hl4BFIvIQsx1ZTvv91EpdZFXk+K0JyJFmJya+8x0InIS0ElVXRGZA/wEY3W/E7i0ocHZvDtahR1L+FhyX0yX1/6iSMHSTZ0HLN3UeQBAgdRUDer+9VKn3/otg3ts6dCpbeVRIhlbtddxse5K36iqe11cRKQUWKqq6cvsB0AXEfm+59o7DpirqttSjhXgDynjz8prwKaYPsZOvsqxsQR3JON8ZqHtZuM4TgkNhHTXaEGrj9Z1OfajdV0AKCqoKR982JaFTr/12wd139qpQ+uqo0TqvWvN6nmf5y3wGt4MJSIXAP/pVX7fUsc4fwbMVNV/eO/vxviJl2NmvwaxKab3McUF/aYAuA34Fwtt+8FlZGGsrKwpaPP+mm6D3l/TDYA2hdW7h/b6aunIvht2HHfo192Ki6oHiOy1er/bWHueENp7v68j5UZAVefRgAesqj6a9v4l4KVMz8VmdEoMcyFugypMYgpb7TcJx3E6Ypz5D/GrzeKiqh3DY5tWjOi7YXe3kj239bxm11y/2vYb2xG9f8deysEnkvHcmp0cx2lSqFOGlAFdXNfN2eRftiN6p1ts+4pYwmqQZ1Y4jnMyxtvRFjNzWUhgX0zP2my83ZYltzqOs58ba9A4jnMo5vFEkcVuXrPYti/YFtMc0kJt/KLz58+9d/jc8ccC0xzHOdhGH5ngOE5bTKyZzTSFuwgmfU+zsCqmZBzF79mpprrq8DkT3u62/LFhYrwOjwZmebNDoDiO0w4TFNlgCJAPPOu6rt+WdN8JIj+Tb/mLCvds+aqvO3ZZ8dbFp6V9dBww23GcwCpDOY7TDVPV8qwAuns4gD6ajXUxJeMsx2S6bRbtNn+8rM+Mi6sKK8vqK0XfH/jYcZyLmttXYziOcxbmgWoQRYQ+cF03LwocBuVsdhdZxq2n0mX5pNkHfz55iDTuz9QReNZxnNHAra7r7vdQtDk4jtMVuAOTDW8/nyFL5MWsBAEmlY8lmIkJOc4Yqa4o7znnl/Pabv+sKdck5Zgv4h7XdZuVw8kzRt4ITMAknw+K1cCRruvuCbDPJhOkmM6AzMs3tNq9cX3snX/dWli1+5hmdl2J8SKcDEx33f1Cq+vEu0tzME/Kz8U4jgXNha7r/iWEfptE0OUuXgV+2Nh+JRvmLuwx/9eHCdrN5yHUYCp3z/N+fu29qjDuGAcDvTDZRk4gXDfhWa7rZjWTh03QDvrXAafTQB6CQ5Y89HbH5LRhYmdsBRiPwiCrDDSFakypi7wi0BJhXiqcO+r6TKr2fBOb9fP3OiWnnWZJSPnEn1zXXRL2ILIljHpzCeCj1A1FO9eu6euOWd1mZ3JYPce0JL6knn+4XCdwMXmpji8DdgN0+HLm/NJZ49oXVJcHUToi16kAxriuu7XRPXOQUGr0AsQSXHrownvHHbR2xqkSfhXzXOEa13X/FPYgmkpoX2IyzlMd185YHAlpL4/ns5Ag/C9yPPC3kMeQCyzABD3mNaGKyXXdakwozbwwxxEyq4Dz88XK3RBhz0x4rhUjMVEQLY1VwOmu665tdM88IHQxAXiPOM6mZS15K4Az/H4YHSY5ISYA13W/wZTZejnssQTAh8Bw13VXN7pnHpEzYgJwXbcCE7H7SNhjscgU4EzXdTeHPRC/Cc3O1BiO41yAKYkVmn+3z+wGbnRdt94yX/lOzooJwHGcw4GnMA+H85lFwEWu634a9kBsklPLXDreXc4I4N+wl43OJgo8CAw50IUEOT4zpeI4zhGYtC4XEpzLbHOYDtzmuu7CsAcSFHkjplocxxmEKQ06qrF9Q2IORkQtzm6Wd2KqxXGc4ZiZKle8ET8E7nZdd1rYAwmLvBVTLd7yNw7j1tI94O5XA88AT7WEa6LGyHsx1eI4TgEmsnYsxs+81FJXG4G/Yu4y33Zd98D4A/rAASOmdBzHOQST2Op7KT+7ZNnMLmAJpnbtPOBd13VXisj1mAKBCvwduFJVN4nIZiD1OdsfVPWZ1Aa9fNwP8W1I+evALapaLSLnYrLdlgNXqOp875hHgedUdUaW4w+UA1ZMdeEFUXbGGEI7p/xeiIlS2Zr22uK6bk1qGyIyGGPFHqiq2716uB0wInhFVRssRiEi44HhwBiMaWY28KCqThaRjzCzagxTDn6Ml7D0JlW1UbvOV1qU4773CKNZjzFUdYGI9FfVShFpi8mj/QUmVLxaRGZjIotfAH6rqtVpx98vIg95ZUu7YSpg1brp1iZ374gpBy/A72kgR2YukdNGy1zFE9L5mCXtVEzp0VaYRBZne9t+QD3Jv7zj78EskRsxsxPALcBzmBKmd2Iqer+pqnnxQLhFLXM2EJErMQlb+3kl3Wu3/xNwg6qmZ2xJPbYI8/xRVPXytM86Ywyfp2E8UocDS1Q141omQRPNTFkiIv1EZHjKpkmYa5xLvZLwe3eljgpMIjJMRI4AM0MBT2AiiNO5G2Oc7QWMUNVRmDzeI3w5EQtEYsqe7sBzItLVe38JsBSTdOxOESkUkXaYTCl1ZXs7E3hARFqJSIF3/D45GERkINBLVafxbWJ3yDC5e2ioavTK8oUxCyzFmAxeA3pjvuRJmDrCKzHltWovI64GHvV+bw38N7AMY3Z4GChOa38GcETK+ycxZS6mAkVhn399r+iaKcI3omUuwjciMUX4RiSmCN+IxBThG5GYInwjElOEb0RiivCNSEwRvhGJKcI3IjFF+EYkpgjfiMQU4RuRmCJ8IxJThG/8P7UGjX0CCExeAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 216x144 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = overlap_counts.loc['DeepBGC'][['Novel','ClusterFinder overlap', 'All overlap', 'antiSMASH overlap']]\n",
    "col = [bin_colors[c] for c in counts.index]\n",
    "total = counts.sum()\n",
    "perc = ['{:.1f}%'.format(c/total*100) for c in counts]\n",
    "ax = counts.plot.pie(colors=col, wedgeprops=dict(width=0.5), labels=perc, figsize=(3, 2), legend=False)\n",
    "ax.set_ylabel('')\n",
    "ax.figure.savefig('../data/figures/paper/main/5x.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 748,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x144 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "counts = overlap_counts.loc['ClusterFinder'][['Novel','DeepBGC overlap', 'All overlap', 'antiSMASH overlap']]\n",
    "col = [bin_colors[c] for c in counts.index]\n",
    "total = counts.sum()\n",
    "perc = ['{:.1f}%'.format(c/total*100) for c in counts]\n",
    "ax = counts.plot.pie(colors=col, wedgeprops=dict(width=0.5), labels=perc, figsize=(3, 2), legend=False)\n",
    "ax.set_ylabel('')\n",
    "ax.figure.savefig('../data/figures/paper/main/5y.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 743,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axes = plt.subplots(1, len(overlap_counts), figsize=(6, 5))\n",
    "i = 0\n",
    "for model, counts in overlap_counts.iterrows():\n",
    "    col = [bin_colors[c] for c in overlap_counts.columns]\n",
    "    total = counts.sum()\n",
    "    perc = ['{:.0f}%'.format(c/total*100) for c in counts]\n",
    "    ax = counts.plot.pie(colors=col, wedgeprops=dict(width=0.5), ax=axes[i], labels=perc)\n",
    "    lgnd = ax.legend(overlap_counts.columns, loc='upper center', bbox_to_anchor=(0.5, -0.1))\n",
    "    if i != 1:\n",
    "        lgnd.remove()\n",
    "    ax.set_ylabel('')\n",
    "    ax.set_xlabel(model)\n",
    "    i += 1\n",
    "#plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 417,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fig.savefig('../data/figures/paper/main/5x.pdf', bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}