Phrank.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "C:\\Users\\levrex\\Desktop\\DeepPhenotypingHPO\n",
      "C:\\Users\\levrex\\Anaconda3\\envs\\my_env\\lib\\site-packages\\clinphen_src\\data\\hpo_synonyms.txt\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2020-12-21 21:15:04,173 - txt2hpo - INFO - txt2hpo 0.2.2\n",
      "2020-12-21 21:15:04,239 - txt2hpo - INFO - Using configuration file: \\Users\\levrex\\.txt2hpo\\txt2hpo.ini\n",
      "2020-12-21 21:15:05,271 - phenopy - INFO - checking if config file exists: \\Users\\levrex\\.phenopy\n",
      "2020-12-21 21:15:05,285 - phenopy - INFO - phenopy 0.4.2\n",
      "2020-12-21 21:15:05,314 - phenopy - INFO - Using configuration file: \\Users\\levrex\\.phenopy\\phenopy.ini\n",
      "2020-12-21 21:15:33,668 - txt2hpo - INFO - Using scispaCy language model\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<module 'DeepPhenotyping_functions' from 'src\\\\DeepPhenotyping_functions.py'>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import sys\n",
    "sys.path.append('src/') # add source with all scripts\n",
    "\n",
    "import pandas as pd\n",
    "import DeepPhenotyping_functions as func\n",
    "from bs4 import BeautifulSoup\n",
    "import importlib as imp\n",
    "imp.reload(func)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "ImportError",
     "evalue": "cannot import name '__version__'",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mImportError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-1-b5137d30abf0>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[1;32mfrom\u001b[0m \u001b[0mphrank\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mphrank\u001b[0m \u001b[1;32mimport\u001b[0m \u001b[0m__version__\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0m__version__\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;31m# alabaster==0.7.10\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mImportError\u001b[0m: cannot import name '__version__'"
     ]
    }
   ],
   "source": [
    "from phrank.phrank import __version__\n",
    "print(__version__) # alabaster==0.7.10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset #1 IARS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "d_IARS_valid = {\n",
    "    'All' : ['validation/021020_1758_HPODP_proj7_art92_pat69-export.html',\n",
    "                            'validation/021020_1758_HPODP_proj7_art92_pat70-export.html',\n",
    "                            'validation/021020_1758_HPODP_proj7_art92_pat71-export.html',\n",
    "                            'validation/021020_1758_HPODP_proj7_art103_pat72-export.html', \n",
    "                            'validation/021020_1758_HPODP_proj7_art103_pat73-export.html', \n",
    "                            'validation/021020_1758_HPODP_proj7_art103_pat74-export.html', \n",
    "                            'validation/021020_1758_HPODP_proj7_art103_pat75-export.html',\n",
    "                            'validation/021020_1757_HPODP_proj7_art113_pat67-export.html',\n",
    "                            'validation/021020_1757_HPODP_proj7_art93_pat68-export.html'\n",
    "                              ],\n",
    "}\n",
    "\n",
    "dataset = 'IARS'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Patient  #65269 (ID: 69)  File Name validation/021020_1758_HPODP_proj7_art92_pat69-export.html\n",
      "Patient  #85880 (ID: 70)  File Name validation/021020_1758_HPODP_proj7_art92_pat70-export.html\n",
      "Patient  #83921 (ID: 71)  File Name validation/021020_1758_HPODP_proj7_art92_pat71-export.html\n",
      "Patient  Family 1-III:4 (ID: 72)  File Name validation/021020_1758_HPODP_proj7_art103_pat72-export.html\n",
      "Patient  Family 1-III:10 (ID: 73)  File Name validation/021020_1758_HPODP_proj7_art103_pat73-export.html\n",
      "Patient  Family 1-III:9 (ID: 74)  File Name validation/021020_1758_HPODP_proj7_art103_pat74-export.html\n",
      "Patient  Family 1-III:11 (ID: 75)  File Name validation/021020_1758_HPODP_proj7_art103_pat75-export.html\n",
      "Patient  Index patient (ID: 67)  File Name validation/021020_1757_HPODP_proj7_art113_pat67-export.html\n",
      "Patient  Index patient (ID: 68)  File Name validation/021020_1757_HPODP_proj7_art93_pat68-export.html\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Patient</th>\n",
       "      <th>ix</th>\n",
       "      <th>HPO-id</th>\n",
       "      <th>Description</th>\n",
       "      <th>assoc genes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>#65269 (ID: 69)</td>\n",
       "      <td>1</td>\n",
       "      <td>HP:0000152</td>\n",
       "      <td>Abnormality of head or neck</td>\n",
       "      <td>2727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>#65269 (ID: 69)</td>\n",
       "      <td>2</td>\n",
       "      <td>HP:0000234</td>\n",
       "      <td>Abnormality of the head</td>\n",
       "      <td>2705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>#65269 (ID: 69)</td>\n",
       "      <td>3</td>\n",
       "      <td>HP:0000929</td>\n",
       "      <td>Abnormality of the skull</td>\n",
       "      <td>1887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>#65269 (ID: 69)</td>\n",
       "      <td>4</td>\n",
       "      <td>HP:0000240</td>\n",
       "      <td>Abnormality of skull size</td>\n",
       "      <td>1267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>#65269 (ID: 69)</td>\n",
       "      <td>5</td>\n",
       "      <td>HP:0040195</td>\n",
       "      <td>Decreased head circumference</td>\n",
       "      <td>1026</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Patient ix      HPO-id                   Description assoc genes\n",
       "0  #65269 (ID: 69)  1  HP:0000152   Abnormality of head or neck        2727\n",
       "1  #65269 (ID: 69)  2  HP:0000234       Abnormality of the head        2705\n",
       "2  #65269 (ID: 69)  3  HP:0000929      Abnormality of the skull        1887\n",
       "3  #65269 (ID: 69)  4  HP:0000240     Abnormality of skull size        1267\n",
       "4  #65269 (ID: 69)  5  HP:0040195  Decreased head circumference        1026"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "case_study = 'All' # 'Orenstein et al, 2016' 'Musante et al, 2017' 'Kopajtich et al, 2016'\n",
    "\n",
    "columns = ['Patient', 'ix', 'HPO-id', 'Description', 'assoc genes']\n",
    "df_valid = pd.DataFrame(columns=columns)\n",
    "\n",
    "for file_name in d_IARS_valid[case_study]:\n",
    "    \n",
    "    with open('Phenotool/' + file_name, 'r') as f:\n",
    "        contents = f.read()\n",
    "\n",
    "        soup = BeautifulSoup(contents, 'lxml')\n",
    "\n",
    "        tables = soup.findAll(\"table\")\n",
    "\n",
    "        cnt = 0\n",
    "        for table in tables:\n",
    "            if table.findParent(\"table\") is None:\n",
    "                cnt += 1\n",
    "        ths = soup.find_all('td')\n",
    "        pat_id = ths[4].text\n",
    "        #pat_id = pat_id.split(' ')[0]\n",
    "        df1 = func.parseTable(table)\n",
    "        df1.columns = ['ix', 'HPO-id', 'Description', 'assoc genes']\n",
    "        #print(df1)\n",
    "        print('Patient ', pat_id, ' File Name', file_name)\n",
    "        df1['Patient'] = [pat_id for i in range(len(df1))]\n",
    "        #print(df1)\n",
    "        df_valid = pd.concat([df_valid, df1])\n",
    "\n",
    "## Ignore entity linking for now!! \n",
    "#df_valid['Patient'] = df_valid['Patient'].apply(lambda x : 'all')\n",
    "df_valid.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Patient</th>\n",
       "      <th>ix</th>\n",
       "      <th>HPO-id</th>\n",
       "      <th>Description</th>\n",
       "      <th>assoc genes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [Patient, ix, HPO-id, Description, assoc genes]\n",
       "Index: []"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_valid[df_valid['Description'].str.contains('ILD')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convert HTML table to PhenoTable\n",
    "Long table needs to be translated to a wide table with HPO codes as header"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>HP:0040195</th>\n",
       "      <th>HP:0008067</th>\n",
       "      <th>HP:0012758</th>\n",
       "      <th>HP:0410042</th>\n",
       "      <th>HP:0031075</th>\n",
       "      <th>HP:0031703</th>\n",
       "      <th>HP:0002492</th>\n",
       "      <th>HP:0000234</th>\n",
       "      <th>HP:0001250</th>\n",
       "      <th>...</th>\n",
       "      <th>HP:0003256</th>\n",
       "      <th>HP:0006579</th>\n",
       "      <th>HP:0002012</th>\n",
       "      <th>HP:0011923</th>\n",
       "      <th>HP:0025032</th>\n",
       "      <th>HP:0004322</th>\n",
       "      <th>HP:0011297</th>\n",
       "      <th>HP:0025031</th>\n",
       "      <th>HP:0000359</th>\n",
       "      <th>HP:0011471</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>#65269 (ID: 69)</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",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>#85880 (ID: 70)</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>#83921 (ID: 71)</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Family 1-III:4 (ID: 72)</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",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</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>1</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>Family 1-III:10 (ID: 73)</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",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Family 1-III:9 (ID: 74)</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",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Family 1-III:11 (ID: 75)</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",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</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>7</th>\n",
       "      <td>Index patient (ID: 67)</td>\n",
       "      <td>1</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</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>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Index patient (ID: 68)</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>9 rows × 279 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                         Id HP:0040195 HP:0008067 HP:0012758 HP:0410042  \\\n",
       "0           #65269 (ID: 69)          1          0          1          0   \n",
       "1           #85880 (ID: 70)          1          0          1          1   \n",
       "2           #83921 (ID: 71)          1          0          0          1   \n",
       "3   Family 1-III:4 (ID: 72)          1          0          1          0   \n",
       "4  Family 1-III:10 (ID: 73)          1          0          1          0   \n",
       "5   Family 1-III:9 (ID: 74)          1          0          1          0   \n",
       "6  Family 1-III:11 (ID: 75)          1          0          1          0   \n",
       "7    Index patient (ID: 67)          1          0          0          1   \n",
       "8    Index patient (ID: 68)          1          1          1          1   \n",
       "\n",
       "  HP:0031075 HP:0031703 HP:0002492 HP:0000234 HP:0001250  ... HP:0003256  \\\n",
       "0          0          0          1          1          0  ...          0   \n",
       "1          0          1          0          1          0  ...          1   \n",
       "2          0          0          0          1          0  ...          0   \n",
       "3          0          0          0          1          1  ...          0   \n",
       "4          0          0          0          1          1  ...          0   \n",
       "5          0          0          0          1          1  ...          0   \n",
       "6          0          0          0          1          1  ...          0   \n",
       "7          0          0          0          1          0  ...          0   \n",
       "8          1          0          0          1          0  ...          0   \n",
       "\n",
       "  HP:0006579 HP:0002012 HP:0011923 HP:0025032 HP:0004322 HP:0011297  \\\n",
       "0          0          0          1          1          0          0   \n",
       "1          0          1          1          0          0          0   \n",
       "2          0          1          1          1          0          0   \n",
       "3          0          0          0          0          1          0   \n",
       "4          0          0          0          0          1          0   \n",
       "5          0          0          0          0          1          0   \n",
       "6          0          0          0          0          0          0   \n",
       "7          0          1          0          1          0          1   \n",
       "8          1          1          0          1          0          1   \n",
       "\n",
       "  HP:0025031 HP:0000359 HP:0011471  \n",
       "0          1          0          1  \n",
       "1          1          1          0  \n",
       "2          1          0          1  \n",
       "3          0          0          0  \n",
       "4          0          0          0  \n",
       "5          0          0          0  \n",
       "6          0          0          0  \n",
       "7          1          0          0  \n",
       "8          1          0          0  \n",
       "\n",
       "[9 rows x 279 columns]"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def generate_pheno_profiles(df_valid):\n",
    "    d_pat = {}\n",
    "    for pat in df_valid['Patient'].unique():\n",
    "        d_pat[pat] = df_valid[df_valid['Patient']==pat]['HPO-id'].unique()\n",
    "    return d_pat\n",
    "\n",
    "d_pat = generate_pheno_profiles(df_valid)\n",
    "df_pheno = func.getPatientProfileTable(d_pat)\n",
    "df_pheno.index = df_pheno.index.set_names(['Id'])\n",
    "df_pheno = df_pheno.reset_index()\n",
    "df_pheno"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## infer HPO"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "import columns_linked_to_hpo\n",
    "\n",
    "\n",
    "def infer_HPO_translate(row, column_to_hpo):\n",
    "    hpo_list = []\n",
    "    cols = list(row.index)\n",
    "    for i in cols:\n",
    "        val = row[i]\n",
    "        #print(val)\n",
    "        if val == 1:\n",
    "            col = i.lower().strip()\n",
    "            if (col in column_to_hpo.keys()) == False :\n",
    "                print(col, col in column_to_hpo.keys())\n",
    "            if col in column_to_hpo.keys():\n",
    "                if type(column_to_hpo[col]) == list:\n",
    "                    hpo_list.extend(column_to_hpo[col])\n",
    "                else :\n",
    "                    hpo_list.append(column_to_hpo[col])\n",
    "    return hpo_list\n",
    "\n",
    "def infer_HPO(row):\n",
    "    hpo_list = []\n",
    "    cols = list(row.index)\n",
    "    for i in cols:\n",
    "        val = row[i]\n",
    "        if val == 1:\n",
    "            hpo_list.append(i)\n",
    "    return hpo_list\n",
    "\n",
    "if dataset == 'IARS':\n",
    "    hp_patients = list(df_pheno.apply(lambda x: infer_HPO(x) , axis= 1))\n",
    "else :\n",
    "    hp_patients = list(df_all.apply(lambda x: infer_HPO_translate(x, columns_linked_to_hpo.column_to_hpo) , axis= 1))\n",
    "#hp_patients"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# CLUSTERING 1 : HRSS clustering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from phenopy import generate_annotated_hpo_network\n",
    "from phenopy.score import Scorer\n",
    "\n",
    "# data directory\n",
    "phenopy_data_directory = os.path.join(os.getenv('HOMEPATH'), '.phenopy\\\\data')\n",
    "\n",
    "# files used in building the annotated HPO network\n",
    "obo_file = os.path.join(phenopy_data_directory, 'hp.obo')\n",
    "disease_to_phenotype_file = os.path.join(phenopy_data_directory, 'phenotype.hpoa')\n",
    "\n",
    "# if you have a custom ages_distribution_file, you can set it here.\n",
    "#ages_distribution_file = os.path.join(phenopy_data_directory, 'xa_age_stats_oct052019.tsv')\n",
    "\n",
    "hpo_network, alt2prim, disease_records = \\\n",
    "    generate_annotated_hpo_network(obo_file,\n",
    "                                   disease_to_phenotype_file,\n",
    "                                   #ages_distribution_file=ages_distribution_file\n",
    "                                   )\n",
    "\n",
    "scorer = Scorer(hpo_network)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import obonet\n",
    "\n",
    "url = 'phenopy_mod/.phenopy/data/hp.obo'\n",
    "graph = obonet.read_obo(url)\n",
    "\n",
    "id_to_name = {id_: data.get('name') for id_, data in graph.nodes(data=True)}\n",
    "name_to_id = {data['name']: id_ for id_, data in graph.nodes(data=True) if 'name' in data}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Assess whether phenotypic specificity is best way foreward"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'name': 'All', 'comment': 'Root of all terms in the Human Phenotype Ontology.', 'xref': ['UMLS:C0444868'], 'ic': 0.0, 'disease_weights': {}, 'depth': 0}\n",
      "IC= 0.0\n",
      "{'name': 'Abnormality of body height', 'def': '\"Deviation from the norm of height with respect to that which is expected according to age and gender norms.\" [HPO:probinson]', 'synonym': ['\"Abnormality of body height\" EXACT layperson []'], 'xref': ['UMLS:C4025901'], 'is_a': ['HP:0001507'], 'created_by': 'peter', 'creation_date': '2008-02-27T02:20:00Z', 'ic': 1.8958170531054654, 'disease_weights': {}, 'depth': 3, 'synonyms': ['Abnormality of body height']}\n",
      "IC= 1.8958170531054654\n"
     ]
    }
   ],
   "source": [
    "cnt = 0\n",
    "for id_, data in hpo_network.nodes(data=True):\n",
    "    #print(data.get('name'))\n",
    "    print(hpo_network.nodes[id_])\n",
    "    if 'ic' in hpo_network.nodes[id_]:\n",
    "        ic = hpo_network.nodes[id_]['ic']\n",
    "        print('IC=', ic)\n",
    "        if ic != 0:\n",
    "            break\n",
    "            #l_synonyms_new.append(str(name_to_id[data.get('name')]) + '\\t' + syn)\n",
    "            #break\n",
    "    cnt +=1\n",
    "    #break\n",
    "    if cnt == 500:\n",
    "        break\n",
    "    \n",
    "    #print(id_)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Positive control"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0\n"
     ]
    }
   ],
   "source": [
    "terms_a = ['HP:0000377']# wide nasal bridge  #hp_patients[0]\n",
    "terms_b = ['HP:0000153']# abnormality #hp_patients[1]\n",
    "\n",
    "print(scorer.score_term_sets_basic(terms_a, terms_b))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Construct distance matrix with HRSS "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9, 'IARS')"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(hp_patients), dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_datasets = ['ACTB', 'ACTB_ACTG1', 'ACTG1']\n",
    "dataset = l_datasets[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\levrex\\Anaconda3\\envs\\my_env\\lib\\site-packages\\numpy\\lib\\function_base.py:380: RuntimeWarning: Mean of empty slice.\n",
      "  avg = a.mean(axis)\n",
      "C:\\Users\\levrex\\Anaconda3\\envs\\my_env\\lib\\site-packages\\numpy\\core\\_methods.py:170: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  ret = ret.dtype.type(ret / rcount)\n"
     ]
    }
   ],
   "source": [
    "l = []\n",
    "for j in range(len(hp_patients)):\n",
    "    for i in range(len(hp_patients)):\n",
    "        l.append(scorer.score_term_sets_basic(hp_patients[i], hp_patients[j]))\n",
    "\n",
    "l_cleaned = [x if str(x) != 'nan' else 0 for x in l ]\n",
    "\n",
    "l2 = []\n",
    "n = len(hp_patients)\n",
    "for j in range(len(hp_patients)):\n",
    "    l2.append(l_cleaned[n*j:n+n*j])\n",
    "df_hrss = pd.DataFrame(l2)\n",
    "df_hrss.to_csv('data/deepdata/distance_matrix_%s_hrss.csv' % (dataset)) # Save distance matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Perform K-means clustering\n",
    "\n",
    "### plain binary metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.spatial.distance import pdist\n",
    "from scipy.spatial.distance import squareform\n",
    "\n",
    "dataset = 'hamming'\n",
    "cols = list(df_all.loc[:, ~df_all.columns.isin(['Category', 'Id', 'Protein', 'Origin'])].columns)\n",
    "\n",
    "bool_dist = ['hamming', 'jaccard', 'dice', 'euclidean']\n",
    "\n",
    "distance = bool_dist[1]\n",
    "\n",
    "df = df_all.copy()\n",
    "pairwise = pd.DataFrame(\n",
    "    squareform(pdist(df[cols], metric=distance)),\n",
    "    columns = df['Id'],\n",
    "    index = df['Id']\n",
    "    \n",
    ")\n",
    "df_hrss = df[cols].copy()\n",
    "X_trans = pairwise.values "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### HRSS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_datasets = ['ACTB', 'ACTB_ACTG1', 'ACTB_ACTG1_NAT6', 'ACTG1', 'IARS']\n",
    "dataset = l_datasets[4]\n",
    "df_hrss = pd.read_csv('data/deepdata/distance_matrix_%s_hrss.csv'  % (dataset), index_col=0)\n",
    "#df_hrss\n",
    "#df_pheno = df_actg1.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Decide nr of clusters with elbow method"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "lalala.py\n"
     ]
    }
   ],
   "source": [
    "import re\n",
    "\n",
    "m = re.search('([^/]*\\.py$)', 'eth/te22ij/lalala.py')\n",
    "print(m.group(1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Nr of clusters = 3\n"
     ]
    }
   ],
   "source": [
    "if dataset in l_datasets:\n",
    "    X_trans = df_hrss[df_hrss.columns[:111]]\n",
    "elif dataset=='IARS':\n",
    "    X_trans = df_hrss\n",
    "k = func.elbowMethod(X_trans, method='kmeans', n=9)\n",
    "print('Nr of clusters = %s' % (k))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Perform K-means"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "import matplotlib.pyplot as plt\n",
    "k =4\n",
    "kmeans = KMeans(n_clusters=k, init='k-means++', max_iter=300, n_init=10, random_state=0)\n",
    "pred_y = kmeans.fit_predict(X_trans)\n",
    "#plt.scatter(X_trans.values[:,0], X_trans.values[:,1])\n",
    "#plt.scatter(kmeans.cluster_centers_[:, 0], kmeans.cluster_centers_[:, 1], s=300, c='red')\n",
    "plt.show()\n",
    "\n",
    "df_hrss['cluster'] = kmeans.fit_predict(X_trans)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[t-SNE] Computing 91 nearest neighbors...\n",
      "[t-SNE] Indexed 111 samples in 0.001s...\n",
      "[t-SNE] Computed neighbors for 111 samples in 0.173s...\n",
      "[t-SNE] Computed conditional probabilities for sample 111 / 111\n",
      "[t-SNE] Mean sigma: 0.689691\n",
      "[t-SNE] KL divergence after 250 iterations with early exaggeration: 53.045925\n",
      "[t-SNE] KL divergence after 1000 iterations: 0.135361\n",
      "KMeans(n_clusters=4, random_state=15)\n",
      "\n",
      "TSNE figure saved under location: TSNE/Kmeans_phenoMap_tsne_All_Jaccard.html\n"
     ]
    }
   ],
   "source": [
    "from bokeh.palettes import Colorblind8\n",
    "from bokeh.palettes import Spectral10\n",
    "#import importlib as imp\n",
    "#imp.reload(func)\n",
    "\n",
    "l_order = ['Gain Of Function (ACTG1)', 'Loss of function (ACTG1)', 'Other (ACTG1)', 'Gain Of Function (ACTB)', 'Loss of function (ACTB)', 'Other (ACTB)']\n",
    "#func.makePCA(X_trans, df_all['Id'],   df_all['Category'], title='_Euclidean', pal=Colorblind8, l_order=l_order, radius=1, seed=2)\n",
    "\n",
    "# X_trans df_hrss[df_hrss.columns[:111]].values\n",
    "func.makeTSNE_Cluster2(X_trans, df_all['Id'],  df_hrss['cluster'].astype(str), df_all['Category'], l_order=l_order,\n",
    "                      title='All_Jaccard', clusters=4, pal=[Colorblind8, Colorblind8], seed=15, color_on_spot=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### alternative UMAP\n",
    "You have to first download umap!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "import umap\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "standard_embedding = umap.UMAP(random_state=42, n_neighbors=30, min_dist=0.3, metric='euclidean').fit_transform(df_hrss[df_hrss.columns[:111]].values)\n",
    "\n",
    "plt.figure(figsize=(15,10))\n",
    "plt.scatter(standard_embedding[:, 0], standard_embedding[:, 1], c= list(pd.Categorical(df_all['Category']).codes), s=100, cmap='Spectral') # c=mnist.target.astype(int),\n",
    "plt.legend(['Gain Of Function (ACTB)', 'Gain Of Function (ACTG1)', 'Loss of function (ACTB)', 'Loss of function (ACTG1)', 'Other (ACTB)', 'Other (ACTG1)'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore Clusters\n",
    "### Plot cluster membership"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2551103c470>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3MAAAHICAYAAADk27PlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+j8jraAAAgAElEQVR4nOzde1zPZ//A8deng0pIx1UOlXNzLLnlVKFiDmEzh8067HaaHZxuhznmcC8ZwjBmrFpmJIfJFqJYP4UwZrfNbGuyMFQYQvX5/ZG+t69v6SDSvffz8fg+8r2u63Nd78+H2+296/BRVFVFCCGEEEIIIUTVolfZAQghhBBCCCGEKDtJ5oQQQgghhBCiCpJkTgghhBBCCCGqIEnmhBBCCCGEEKIKkmROCCGEEEIIIaogSeaEEEIIIYQQogoyqOwASmJlZaU6OjpWdhhCCCGEEEIIUSmOHTt2VVVV60fLn/tkztHRkdTU1MoOQwghhBBCCCEqhaIovxdVLssshRBCCCGEEKIKkmROCCGEEEIIIaogSeaEEEIIIYQQogqSZE4IIYQQQgghqiBJ5oQQQgghhBCiCpJkTgghhBBCCCGqoOf+1QRCCCGEEP8rcnJyuHLlCjk5OeTm5lZ2OEKISmZoaIiNjQ21atUq1/WSzAkhhBBCPAPXr1/n8uXLWFtbY2tri4GBAYqiVHZYQohKoqoqd+7c4Y8//gAoV0InyyyFEEIIIZ6Bq1evUrduXczNzTE0NJREToi/OUVRqF69OnXq1OHPP/8sVx+SzAkhhBBCPAP37t3DxMSkssMQQjxnTExMuH//frmuLXcypyjKG4qiqA8+w4tp01FRlK8VRclUFOW2oiinFEUZpyiKfnnHFUIIIYSoqmQ2TgjxqCf5e6FcyZyiKPWAj4C/HtOmH3AQ8AC2ASuBakAY8GV5xhVCCCGEEEIIUaDMyZxSkDp+BlwDVhfTphawFsgDvFRV/aeqqpOANkAyMFBRlCHljloIIYQQQggh/ubKc5rle0A3wOvBz6IMBKyBSFVVUwsLVVXNURRlBrAPeAuZoRNCCCGEwHHqrkodP21B7ye6Pjk5mWXLlpGUlMSff/6JsbExTZs2pWfPnowZMwY7O7sy9ZeYmEjXrl1JSEjAy8vriWJ72NGjR1mwYAFJSUlkZ2djZ2dHr169mD59OnXq1NFqe+vWLUaPHs3u3bu5cuUKY8eOZenSpUX2W9wyuddff52oqKgKi78stm/fzq+//sqECRO0yp/Wsy2tvn374ujoyEcffaRTN3/+fGbOnMmAAQPYunVrsX188803rFy5kiNHjpCVlYWFhQXt27cnKCiIAQMGlGrZooODA2lpaQQHBzNnzhxNub6+PnXq1MHb25u5c+dq/bkYO3Ys586dY9euyv3f68PKlMwpiuIMLACWqap6UFGU4pK5wvK4IuoOAreBjoqiGKmqercsMQghhBBCiOfH4sWLmTRpEl27dmX+/Pk0aNCAv/76i0OHDvHJJ5+QmprKN998U6Y+XV1dSU5O5sUXX6ywOD///HOCgoLo3Lkzy5Ytw97enjNnzrBw4UK2bNlCfHw8rVq10rRfuXIlGzduZP369TRp0qTEhDQwMJBRo0ZplVlbW1dY/GW1fft24uPjdZK5p/FsS+vgwYPs3buXX375pcj6yMhIAHbt2sW1a9ewtLTUaTNx4kSWLFnCwIEDWbFiBba2tly+fJmdO3cyaNAgUlNTSU5O1rpmwIABtG7dmuDgYE2ZkZGRVpukpCT09fW5f/8+//nPf5g9ezbHjh3j+PHj6OkVLGacOnUqDRo0YP/+/XTrVlwa9GyVOplTFMUA+Bw4D0wroXnTBz/PPlqhqmquoii/Ac2BBsCZ0sYghBBCCCGeHwkJCUyaNImxY8cSFhamVderVy/ef/99oqOjy9xvrVq1cHd3r6gw+emnnxgxYgT9+/dn8+bNmn+ce3h4MHDgQNq3b8/AgQP54YcfMDQ0BODMmTPY29vj7+9fqjHq1KlToTE/LRX9bMviww8/pG/fvjqzoACHDh3i559/plevXnz99dds3LiRd955R6tNVFQUS5YsYdGiRUycOFGr7tVXX2Xs2LGYm5vTunVrrTojIyOsrKwee9/t27fHwKAgNerSpQv6+vqMGDGCn376CWdnZwDs7Ozo27cvixYtem6SubLsmZsFuACBqqreKaGt2YOf14upLyyvXVSloigjFUVJVRQl9cqVK2UI8W8i2KxiPkIIIYQQTyA0NBQrKytCQ0OLrDc1NSUwMFCrbPbs2bi6umJmZoaVlRXdunUjJSVFq01iYiKKopCYmKgp8/LyonPnzsTHx+Pq6kr16tVp0aIF27dvLzHOpUuXkpeXx0cffaRJ5ApZWlrywQcf8PPPP2uW9imKQnh4OOnp6SiKohNLWTk6Ouo8h8JxHp4tCg4ORlEUfv75Z3r37k2NGjVwcHBg7ty55Ofna1175coVxowZQ7169TAyMqJevXq88cYb3L17l8DAQCIiIvjjjz808Ts6OgJFP1tVVQkLC6Np06ZUq1YNOzs73nnnHW7cuKET74wZM1i+fDlOTk7UrFkTT09PfvjhhxKfQUZGBt988w2vvfZakfURERHo6+uzdu1a6tWrp5mle9gHH3xAixYtdBK5Qm3btqV+/folxlIahS/wfvSVAUOGDGH37t2kp6dXyDhPqlTJnKIo/6BgNm6xqqrJJbUvTZcPfqpFVaqq+omqqm6qqrpV5vS0EEIIIYQoWm5uLgcOHMDHx4dq1aqV+ro//viD8ePHs337dsLDw7GxscHDw4NTp06VeO0vv/zC2LFjmTBhAlu3bsXOzo6BAwdy7ty5x163b98+3Nzcil0q2bt3b/T09Ni/fz9QsAewR48e2NrakpycTHJyMq6uro8dQ1VVcnNztT7lNWDAALp168b27dvp378/s2fPJiIiQlOflZVFx44d2bRpExMmTODrr79m4cKF3L9/n3v37jFz5kx69eqFtbW1Jv5t27YVO9706dOZMGECPj4+7Ny5k8mTJxMeHk7v3r11ksioqCh27drFsmXL+Oyzzzh//jz9+vUr8X737t1LXl4enTt31qnLyclh8+bN+Pj4YG9vz7Bhwzh69Chnzvx3AV9GRgZnzpyhb9++pX2MZZKXl0dubi537tzh2LFjfPDBBzRv3pwWLVpotfPw8CA/P5+9e/c+lTjKqsRllg8trzwLzCxlv4Uzb8VN/9R6pJ0QQgghhKhCrl27Rk5OTpEzIY/+w75w+RrAp59+qvl1Xl4ePXv2pHnz5qxbt45ly5Y9dsyrV69y8OBBGjduDBTs/7Kzs2Pz5s1Mm1b8LqD09HTatm1bbL2pqSnW1taa2RZ3d3esrKwwMjIq9ZLEDz74gA8++ECr7Oeff6ZRo0aluv5hEydOJCgoCABvb2/279/Pxo0bNWVhYWH8+uuvpKam4uLiorlu6NChANSsWRNra2uqVatWYvyZmZksWbKEgIAAVqxYAUCPHj2wtrbmjTfeIDY2Fj8/P017Q0NDYmNjNctRoWCJ45EjR+jYsWOx46SkpGBvb1/kPsIdO3aQnZ2tWdIaEBBASEgIERERLFiwAEDze+Pg4PDY+ykvY2Njre/NmjUjNjZWZybXysqKunXrkpKSwptvvvlUYimL0szM1QCaAM5AzkMvCleB2Q/arH1QVnjEz08PfjZ5tLMHyaETkAv8+kTRCyGEEEKISqGqRS6w4tKlSxgaGmp9Hk7u4uPj6dq1K5aWlhgYGGBoaMjZs2f56aefiuzvYY0bN9YkcgA2NjbY2Nhw/vz5p3Y/pfXmm29y9OhRrU+9evXK1Vfv3tqni7Zo0ULrHvfs2UO7du20ErnySklJ4e7duwwbNkyrfMiQIRgYGHDgwAGtch8fH61ErmXLlgAl/h5kZGQUeyBMREQEtWrVon///gA0bdqU9u3bExUVpTMz+LSkpKRw9OhRDh8+zObNmzE1NcXX15fLly/rtLW2tiYjI+OZxFWS0hyAchdYV0ydKwX76JIoSOAKl2DuB14HegIbH7nGA6gOHJSTLIUQQgghqiYrKyuMjY11/hFvZWXF0aNHAfjkk09Yu3atpu748eP06tWLHj16sG7dOuzs7NDX12f48OHk5OSUOKaFhYVOmZGRUYnX1q1bl7S0tGLrb926xdWrV8udfEHB4Rhubm7lvv5hj97no/d47do1nUM+yiszMxNAZwmqgYEBlpaWmvrHxQaU+HuQk5Ojc4IkFCT/e/bsYdCgQdy9e5e7dwvSg1deeYXJkyezb98+fHx8NL83v//+exnurvTatm2rmUH+xz/+gYeHB3Z2dixZskRnT6iJiQl37pR0hMizUeLMnKqqd1RVHV7UB/jqQbOIB2WbHnzfAlwFhiiKovlTrSiKMTD/wdePK/A+hBBCCCHEM2RgYICHhwd79+7l3r17WuVubm64ublhb2+vdU1MTAwGBgZs3bqV/v370759e9zc3MjKynqqsXbv3p3U1FQuXrxYZP2uXbvIz89/aicUGhsbaz0jQCdJKgsrKyv++OOPJw0L+G9ydunSJa3y3NzcYl8PUB6WlpZF/j5HRUWRl5fHxo0bMTc313wmT54MoNkraG9vj7OzMzt37qyQeErywgsvYGVlVeRezszMTKysrJ5JHCUpy2mWpaaq6g1gBKAPJCqK8qmiKAuB74AOFCR7mx7ThRBCCCGEeM5NnjyZq1evMmXKlFK1v337Nvr6+lovdd6/f3+FLJN8nLFjx6Knp8e7776rs2wvMzOTadOm0ahRI15++eWnMr6DgwOnT5/WKouNjS13f76+vhw5coSTJ08W28bIyKhUs0fu7u4YGRnx5ZdfapVv2rSJ3NxcPD09yx3nw5o1a0Z6errOfsrIyEgcHBxISEjQ+fTs2ZNt27Zx8+ZNAKZNm8bp06dZsmRJkWOcOHGiwv4sXbx4katXr+osDc3LyyM9PZ2mTZsWc+WzVaaXhpeFqqrbFUXxBKYDrwDGwDlgArBcfdKFyUIIIYQQolJ1796dBQsWMHXqVE6dOoW/vz9OTk7k5ORw9uxZvvzyS0xNTTXJW8+ePVm6dCmBgYEEBQVx9uxZ5s2bV+R7xyqSs7Mza9asYfjw4XTv3p3Ro0djZ2fHjz/+yMKFC8nOzmbv3r1ae8Eq0pAhQ3jzzTcZP348ffr04eTJk4SHh5e7v/Hjx/PFF1/g7e3NjBkzaNmyJVevXmXHjh2sXr2amjVr8uKLL5KZmcnHH3+Mm5sbxsbGmv1tD7OwsGDChAmEhIRgampKr169OHPmDDNmzKBz5846+/fKy8PDg9mzZ3Pq1CnNyaDHjx/n+++/Jzg4GC8vL51rcnJyiIuLY8uWLQQFBTFs2DCOHz/OxIkTSU5OZtCgQdja2vLnn3+ya9cuPv/8c1JTU8v1eoLDhw+jr69Pfn4+v//+Ox9++CH6+vqMHj1aq93p06e5desWHh4e5XoOFe2JkjlVVYOB4MfU/x/Q60nGEEIIIYT4X5e2oGL+wVwZJk+eTKdOnVi2bBnTpk3jypUrGBsb07RpUwYPHszo0aPR19cHCk5JXL58OUuWLCEmJoYWLVoQGRnJ/PnzSxjlyQUGBtKsWTNCQ0N55513uH79OnZ2drz00ktMnz79ifbLlSQgIID09HTWrVvHmjVr6NKlC9u2bSvXSZcAtWvX5v/+7/+YMWMGCxYs4Nq1a7zwwgt069ZN85qI4cOHk5KSwrRp08jOzsbBwaHYfYP//ve/sba2ZvXq1axatQpLS0v8/f0JCQnROc2xvLp06YK9vT07d+7UJHMREREoilLkO/igYAayXr16REREaE7yXLJkCd7e3qxcuZIxY8aQnZ2NhYUF7u7ubN26tdx7CQtfmaAoCra2trRt25bVq1fzj3/8Q6tdbGwstra2RSaflUF53ifI3Nzc1NTU1MoO4/lSUS/8DpY3QwghhBDPypkzZ3B2dq7sMISoNMHBwWzYsIGzZ89qLbWtSl588UVeeeUV5s2bV6H9lvT3g6Iox1RV1Tlh56nsmRNCCCGEEEKIh40fP57s7GxiYmIqO5Ry2bFjB5cvX2bixImVHYqGJHNCCCGEEEKIp87MzIzPP/9c52TPquLOnTtERUVRu3btyg5F46kdgCKEEEIIIYQQD+vZs2dlh1BuQ4YMqewQdMjMnBBCCCGEEEJUQZLMCSGEEEIIIUQVJMmcEEIIIYQQQlRBkswJIYQQQgghRBUkyZwQQgghhBBCVEGSzAkhhBBCCCFEFSTJnBBCCCGEEEJUQfKeOSGEEEKIyhZsVsnjXy/XZeHh4QQFBfHzzz/TqFGjCg7q2cnPz2fChAls2rSJy5cv4+fnx/bt24tse+nSJUaOHElSUhJZWVmEhYUxbty4ZxwxpKWlER4ejr+/Pw0aNNCqc3R0xMvLi/Dw8GceV0xMDGPGjOG3336jevXqWnXp6ek4Ojqir69PRkYGVlZWRfZx/vx5Fi5cyO7du0lPT8fAwABnZ2f69+/PO++8w9ixY4mIiCgxloSEBBwdHXFyctIqt7S0pHnz5kybNo0ePXpoyk+cOEGnTp348ccfqV+/fjnu/tmTZE4IIYQQQvytbdmyhWXLlrF48WI6dOiApaVlsW3nzp3LgQMHCA8Px87ODkdHx2cX6EPS0tKYM2cOnTt31knmtm3bRq1atZ55TLm5uUybNo1JkybpJHIAkZGR5Ofnk5+fz8aNG3n33Xd12hw8eBA/Pz9sbGx47733aNGiBffv3yclJYWVK1dy9epVZs6cyejRozXXfPrpp6xbt46kpCT09fU15S+++CKZmZkAvP/++/j5+QFw+fJlVqxYQZ8+ffj2229xd3cHwMXFBR8fH2bOnFmqZPF5IMmcEEIIIYT4Wztz5gwA48aNQ0/v8buQzpw5Q+vWrRkwYMCzCK1cXFxcKmXcHTt2kJaWxptvvllkfWRkJC1atODGjRtEREToJHNZWVkMHDgQZ2dn4uPjMTU11dT5+voyceJEDh06RMOGDWnYsKGmLi4uDoD27dtjYKCd3hQmcw0aNNAkbQA+Pj6Ym5uzbds2rfJRo0bRr18/QkJCsLe3L+eTeHZkz5wQQgghhHhq7t+/z4wZM3B0dKRatWo4OjoyY8YM7t+/r2mTm5vLzJkzadiwIcbGxlhZWdG5c2eSkpI0bb744gtcXFyoUaMGZmZmtGzZkjVr1pQ4flxcHB06dMDExAQzMzP69+/PTz/9pKl3dHQkODgYAH19fRRFKXJ5YlpaGoqikJiYyLfffouiKCiKQlpaGsHBwSiKonNNYGCg1sxdYR9r1qxh1qxZ2NnZUbt2bfr27cuFCxd0rl+7di2urq6YmJhgbm6Op6cnhw4dIjExka5duwIFSUlhLImJiZp7CgwM1OrryJEjeHt7U6NGDUxNTenevTtHjhzRibdu3bqcOHGCLl26UL16dRo3bszq1atLfM5QMEPWs2dPLCwsdOqSk5M5e/Ys/v7+vPHGGxw7dowffvhB536vXLnCRx99pJXIFTI1NcXHx6dUsZTE2NgYQ0NDrT+HUJA01qpVq1KWqJaHJHNCCCGEEOKpCQgIYMGCBfj7+xMbG0tQUBChoaEEBARo2oSGhhIWFsZ7773H7t27+eyzz+jevbtmViUpKYlhw4bh6enJ9u3biY6OZsSIEWRnZz927Li4OHr37k2NGjXYtGkTH3/8MadPn6Zz58788ccfQMGSxMLEJzk5meTkZHr37q3Tl52dHcnJybRq1QoXFxdNWzs7uzI/k5CQEM6dO8f69etZtmwZycnJvP7661pt/vWvfzFy5EhcXV3ZvHkzUVFReHh4cP78eVxdXVm5ciUAy5cv18Ti6upa5HinTp3C09OTrKwswsPDiYyM5MaNG3h6enLy5Emttjdu3OC1115j2LBh7Nixg3bt2vHWW2+RkJDw2Hu6e/cuiYmJdOnSpcj6iIgI9PT0eP311/H39wcKZuoeFh8fj62tLW5ubo8dqzzy8/PJzc0lNzeXjIwMJk6cSE5ODq+88opWOwMDAzp06KCZ7XveyTJLIYQQQgjxVJw+fZqNGzcye/ZszeyXr68v+vr6zJw5k6lTp9KqVSuSk5Px9fVl7Nixmmv79u2r+XVKSgq1a9dm6dKlmjJfX98Sx58xYwYNGjTgm2++0Sy/69ChA02aNGHx4sUsWbIEFxcX6tSpA6C13O5RRkZGuLu7U7NmTQwMDB7btiQODg588cUXmu9Xrlxh0qRJZGRkYG9vz7lz5wgLC2P8+PEsWbJE0+7hJPPFF18EwNnZucRY5s6di5GREfv27aN27dpAwYyeo6Mjc+bMYevWrZq2N2/eZNWqVZqZPw8PD/bs2cPGjRs1ZUX57rvvyMnJoXXr1jp1d+/eZdOmTXh7e2uWLrq7uxMVFUVISIhmaWvhASlPw6hRoxg1apTmu5GREZ988gmdOnXSaevi4sKHH35Ifn5+ictuK9vzHZ0QQgghhKiyDh48CMCwYcO0ygu/HzhwAIB27drx9ddfM336dJKSkrh3755W+3bt2pGVlcWwYcOIjY0tcUYO4NatWxw/fpzBgwdr7aNycnKiU6dOmrErw6Mzfy1btgQKTnGEghmq/Px8Ro4cWSHjHTx4kD59+mgSOYBatWrh5+en8xyqV6+ulbQZGRnRuHFjTWzFycjIAMDa2lqnbseOHWRnZ2tm5KBgxjYjI4P4+Phy3VNZzZgxg6NHj3L06FF2797NiBEjGDlyJF9++aVOW2tra+7evauZGX6eSTInhBBCCCGeisJ/DD+6FNHW1larftq0acyZM4evvvqKLl26YGlpSVBQEFevXgXA09OT6Oho0tPTGTBgANbW1nh7e3Pq1Klix87KykJV1SKXQdra2lbqP9Qf3VNmZGQEQE5ODgDXrl0DoG7duhUyXmZmZrHPISsrS6vM3Nxcp52RkZEmtuIU1hfey8MiIiI0SWJ2djbZ2dn06NEDQ0NDrVMj69WrR1paWmluqcwcHBxwc3PDzc0NX19fPvroI3r06MG4ceNQVVWrrYmJCQB37tx5KrFUJEnmhBBCCCHEU1GYtFy6dEmrvPB74SsADA0NmTJlCt9//z0XL14kLCyMmJgY3n77bc01AwcO5MCBA2RlZbFt2zYuXrxIz549yc/PL3Jsc3NzFEXRGbtw/Me9fqCsjI2NAXRmFAuTsrIqfP9a4b6+J2VhYVHscyjqsJLyKHyejyaHly9fZs+ePdy+fZs6depgbm6Oubk5DRo04P79+2zbto2bN28C4O3tzaVLlzh27FiFxFSS5s2bc/nyZf7880+t8sJEv7j34D1PJJkTQgghhBBPhaenJ4DOUrYNGzYABfuxHmVra8vw4cPx9vbm9OnTOvU1atSgT58+jBo1iosXLxabMJmamtK2bVuio6PJy8vTlP/+++8cOnRIE1tFcHBwANCKNzs7m0OHDpWrP29vb/T09Pjkk0+KbVM4A1aa2SNPT0927dqlSZqgYG/czp07K+w5NGvWDIBff/1VqzwqKorc3Fw+/vhjEhIStD5Lly7lzp07REdHAzB8+HCsrKx45513uHXrls4Yt2/frtBlmadOnaJatWqYmZlplf/222/Uq1dPM0P3PJMDUIQQQgghxBOJi4vTLJ0sZGZmho+PD0OHDiU4OJjc3Fw6duxIcnIy8+bNY+jQobRq1QqAfv360bp1a1xdXTE3N+fEiRPExcVpDqyYNWsWly9fpmvXrtjb23PhwgWWL19OmzZtityjVWjevHn07t2bPn36MGbMGP766y9mz56NmZkZEydOrLD7f+mllzAzM2PEiBHMmTOHu3fvsnDhQmrUqFGu/ho2bKg5/OTmzZv4+fmhr6/PkSNHaNasGYMHD6ZJkyYYGBiwfv16LCwsMDIyomnTptSsWVOnv5kzZxIbG0v37t2ZMmUKiqIQGhrK7du3mTVr1pPePgD169fHwcGBI0eOaO2RjIyMxMnJiVGjRum8vqFLly4sXLiQiIgI3nzzTSwsLIiJicHPzw9XV1feffddzUvDjxw5wurVqxk4cCDe3t5lju/XX38lJSUFKJg93LlzJ7t372bMmDGamdVChw8fLvI/NDyPJJkTQgghhKhswdcrO4In8ujLn6FgCdvp06eJiIigQYMGrF+/nvnz52Nvb8+UKVOYPXu2pq2HhwfR0dGsXLmS27dvU79+fSZPnsz06dOBgpdBL1++nPHjx5OZmYmNjQ2+vr7MmzfvsXH17NmTXbt2MWfOHAYNGkS1atXw8vJi4cKFFfpC6Nq1axMbG8v48eMZNGgQdevWZdasWcTHx2ve/VZWixYtolGjRqxatYqIiAhMTU1p1aqV5hRPS0tLVqxYQWhoKJ6enuTl5ZGQkICXl5dOX61atSIxMZHp06cTEBCAqqq4u7tz4MCBIk+fLK/BgwcTHR3N8uXLAThx4gSnTp1i7ty5Rb6HT19fn8DAQEJCQvjtt99wcnLCw8ODkydP8uGHHxIWFsaFCxcwNDTE2dmZt99+mzFjxpQrtpCQEEJCQoCCw18aNmzIypUrdQ6ZSU9P5+TJkyX+2XpeKI9u+HveuLm5qampqZUdxvMl2KzkNqXqp2r/H4cQQghRlZw5cwZnZ+fKDkOIp+aXX36hadOmJCYm0rlz58oOp1xCQ0P5+OOP+eWXX9DX139m45b094OiKMdUVdV5AZ/smRNCCCGEEEI8sYYNGxIUFMSCBQsqO5RyycnJYdmyZcydO/eZJnJPQpI5IYQQQgghRIWYN28e7dq14/bt25UdSpmlpaUxduxY3njjjcoOpdRkz5wQQgghhBCiQtja2mrth6xKmjVrpjmVs6qQmTkhhBBCCCGEqIIkmRNCCCGEEEKIKkiSOSGEEEIIIYSogkqdzCmKEqooyj5FUdIVRbmjKEqmoignFEWZrSiK5SNtHRVFUR/z+bLib0UIIYQQQggh/j7KcgDKeOA4sBf4EzAF3IFgYKSiKO6qqqY/cs1JYHsRfZ0ue6hCCCGEEEIIIQqVJZmrpapqzqOFiqL8G5gGvA88+kr271RVDS5/eEIIIYQQQgghilLqZZZFJXIPbH7ws/GThyOEEEIIIYQQojQq4j1zfUTXP8cAACAASURBVB/8PFVEnb2iKKMAS+AakKyqalHthBBCCCH+tlpGtKzU8b8P+L5c14WHhxMUFMTPP/9Mo0aNKjiqZyc/P58JEyawadMmLl++jJ+fH9u3F7VTCC5dusTIkSNJSkoiKyuLsLAwxo0b94wjLnjBdXh4OP7+/jRo0ECrztHRES8vL8LDw595XDExMYwZM4bffvuN6tWra9Wlp6fj6OiIvr4+GRkZWFlZFdnH+fPnWbhwIbt37yY9PR0DAwOcnZ3p378/77zzDmPHjiUiIqLEWBISEvDy8tL0uWjRInbv3s358+dRFAUHBwc8PT0ZOXIkrq6umut27tzJxo0bSU1N5dy5c3h4eJCYmKjTf79+/ahbty4rV64s/QOqYGVO5hRF+RdQAzAD3IDOFCRyC4po7vPg8/D1iUCAqqrnyzq2EEIIIYQQFW3Lli0sW7aMxYsX06FDBywtLYttO3fuXA4cOEB4eDh2dnY4Ojo+u0AfkpaWxpw5c+jcubNOMrdt2zZq1ar1zGPKzc1l2rRpTJo0SSeRA4iMjCQ/P5/8/Hw2btzIu+++q9Pm4MGD+Pn5YWNjw3vvvUeLFi24f/8+KSkprFy5kqtXrzJz5kxGjx6tuebTTz9l3bp1JCUloa+vryl/8cUXAUhMTKRfv3688MILjBkzhlatWqGqKqdOnSI8PJyIiAju3LmjuW779u189913uLu7k5NT3OJECA4Opn379owdO5YmTZqU65k9qfLMzP0LeOGh73FAoKqqVx4quw3Mo+Dwk18flLWi4LCUrsA+RVHaqKp6q6gBFEUZCYwEqF+/fjlCFEIIIYQQonTOnDkDwLhx49DTe/wupDNnztC6dWsGDBjwLEIrFxcXl0oZd8eOHaSlpfHmm28WWR8ZGUmLFi24ceMGEREROslcVlYWAwcOxNnZmfj4eExNTTV1vr6+TJw4kUOHDtGwYUMaNmyoqYuLiwOgffv2GBhopzfXrl1j4MCBtGjRgr1792olmd27d2fcuHF8/PHHWtesXbtW8+egc+fOxd6vi4sLbdq0YenSpaxatepxj+apKfN75lRVtVVVVQFsgZeBBsAJRVFcH2rzp6qqs1RVPa6qavaDz0HAFzgMNAKGP2aMT1RVdVNV1c3a2rqsIQohhBBCiOfE/fv3mTFjBo6OjlSrVg1HR0dmzJjB/fv3NW1yc3OZOXMmDRs2xNjYGCsrKzp37kxSUpKmzRdffIGLiws1atTAzMyMli1bsmbNmhLHj4uLo0OHDpiYmGBmZkb//v356aefNPWOjo4EBwcDoK+vj6IoRS5PTEtLQ1EUEhMT+fbbb1EUBUVRSEtLIzg4GEVRdK4JDAzUmrkr7GPNmjXMmjULOzs7ateuTd++fblw4YLO9WvXrsXV1RUTExPMzc3x9PTk0KFDJCYm0rVrVwB8fHw0sRQuBXR0dCQwMFCrryNHjuDt7U2NGjUwNTWle/fuHDlyRCfeunXrcuLECbp06UL16tVp3Lgxq1evLvE5Q8EMWc+ePbGwsNCpS05O5uzZs/j7+/PGG29w7NgxfvjhB537vXLlCh999JFWIlfI1NQUHx8fnfLHWbt2LdeuXWPFihVFzhYqisKYMdpnOJaU0D9syJAhbNiwQWtm71kq90vDVVW9rKrqNgoSNEsgshTX5AKfPvjqUd6xhRBCCCFE1RAQEMCCBQvw9/cnNjaWoKAgQkNDCQgI0LQJDQ0lLCyM9957j927d/PZZ5/RvXt3MjMzAUhKSmLYsGF4enqyfft2oqOjGTFiBNnZ2Y8dOy4ujt69e1OjRg02bdrExx9/zOnTp+ncuTN//PEHULAksTDxSU5OJjk5md69e+v0ZWdnR3JyMq1atcLFxUXT1s7OrszPJCQkhHPnzrF+/XqWLVtGcnIyr7/+ulabf/3rX5q9XJs3byYqKgoPDw/Onz+Pq6urZp/W8uXLNbE8vO/rYadOncLT05OsrCzCw8OJjIzkxo0beHp6cvLkSa22N27c4LXXXmPYsGHs2LGDdu3a8dZbb5GQkPDYe7p79y6JiYl06dKlyPqIiAj09PR4/fXX8ff3Bwpm6h4WHx+Pra0tbm5ujx2rLPbt24ednd1Tm6308PDgxo0bJCcnP5X+S/LEB6Coqvq7oij/AdooimKlqurVEi4pXI6pm24LIYQQQoj/GadPn2bjxo3Mnj1bM/vl6+uLvr4+M2fOZOrUqbRq1Yrk5GR8fX0ZO3as5tq+fftqfp2SkkLt2rVZunSppszX17fE8WfMmEGDBg345ptvNMvvOnToQJMmTVi8eDFLlizBxcWFOnXqAODu7l5sX0ZGRri7u1OzZk0MDAwe27YkDg4OfPHFF5rvV65cYdKkSWRkZGBvb8+5c+cICwtj/PjxLFmyRNPu4SSzcD+Ys7NzibHMnTsXIyMj9u3bR+3atYGCGT1HR0fmzJnD1q1bNW1v3rzJqlWrNDN/Hh4e7Nmzh40bN2rKivLdd9+Rk5ND69atderu3r3Lpk2b8Pb2xt7eHih41lFRUYSEhGhmwgoPSKlIFy5cwMHBQac8Ly8PVVU13wtnZcuqdevW6OnpkZKSQrdu3Z4o1vIo98zcI+wf/MwrRdvCP22/PraVEEIIIYSo0g4ePAjAsGHDtMoLvx84cACAdu3a8fXXXzN9+nSSkpK4d++eVvt27dqRlZXFsGHDiI2NLXFGDuDWrVscP36cwYMHa+2jcnJyolOnTpqxK8OjM38tWxacZnr+fMH5gPHx8eTn5zNy5MgKGe/gwYP06dNHk8gB1KpVCz8/P53nUL16da2kzcjIiMaNG2tiK05GRgYARW2R2rFjB9nZ2ZoZOSiYsc3IyCA+Pr5c91RaDydsD2vevDmGhoaaz759+8rVv6GhIWZmZpr7f9ZKlcwpitJMURTbIsr1Hrw03AY4pKpq1oPy9oqiVCuifTdg/IOvUeUPWwghhBBCPO8Kl0k+uhTR1tZWq37atGnMmTOHr776ii5dumBpaUlQUBBXrxYs+PL09CQ6Opr09HQGDBiAtbU13t7enDpV/BuvsrKyUFW1yGWQtra2mrErw6N7yoyMjAA0Jydeu3YNgLp161bIeJmZmcU+h6ysLK0yc3NznXZGRkaPPdUR/ht74b08LCIiQpMkZmdnk52dTY8ePTA0NNR6xUC9evVIS0srzS2VWr169YpMRGNiYjh69Gip9wM+jomJyXO/Z64nkK4oyj5FUT5RFCVEUZT1wM/ANOASMOKh9qHAH4qiRCuKEvbgsw/YBxgBM1VVPVSB9yGEEEIIIZ4zhUnLpUuXtMoLvxe+AsDQ0JApU6bw/fffc/HiRcLCwoiJieHtt9/WXDNw4EAOHDhAVlYW27Zt4+LFi/Ts2ZP8/PwixzY3N0dRFJ2xC8d/3OsHysrY2BhAZ0axMCkrq8L3rxXu63tSFhYWxT6Hog4rKY/C5/locnj58mX27NnD7du3qVOnDubm5pibm9OgQQPu37/Ptm3buHnzJgDe3t5cunSJY8eOVUhMAN26dSMjI4MTJ05olTdv3hw3NzeaNm36xGNkZmYW+868p620yVw88AkFB528DEwCXgEygTlAc1VV//NQ+88pOLWyHQVJ3higMbAZ8FBVdX6FRC+EEEIIIZ5bnp6eAHz55Zda5Rs2bAAK9mM9ytbWluHDh+Pt7c3p06d16mvUqEGfPn0YNWoUFy9eLDZhMjU1pW3btkRHR5OX99+dQL///juHDh3SxFYRCvdkPRxvdnY2hw6Vb+7C29sbPT09Pvnkk2LbFM6AlWZGyNPTk127dmmSJijYG7dz584Kew7NmjUD4NdftXdSRUVFkZuby8cff0xCQoLWZ+nSpdy5c4fo6GgAhg8fjpWVFe+88w63bum+wez27dtlXpY5YsQIzM3Neffdd7l9+3Y57654ly5dIicnp0KSwvIo1QEoqqqeBt4useF/268D1pU3KCGEEEIIUXXExcVplk4WMjMzw8fHh6FDhxIcHExubi4dO3YkOTmZefPmMXToUFq1agVAv379aN26Na6urpibm3PixAni4uIYNWoUALNmzeLy5ct07doVe3t7Lly4wPLly2nTpk2Re7QKzZs3j969e9OnTx/GjBnDX3/9xezZszEzM2PixIkVdv8vvfQSZmZmjBgxgjlz5nD37l0WLlxIjRo1ytVfw4YNNYef3Lx5Ez8/P/T19Tly5AjNmjVj8ODBNGnSBAMDA9avX4+FhQVGRkY0bdqUmjVr6vQ3c+ZMYmNj6d69O1OmTEFRFEJDQ7l9+zazZs160tsHCt4N7eDgwJEjR7T2SEZGRuLk5MSoUaN0Dhjp0qULCxcuJCIigjfffBMLCwtiYmLw8/PD1dWVd999V/PS8CNHjrB69WoGDhyIt7d3qeOysrIiOjqaAQMG0KZNG81Lw/X09EhPTycyMhJFUbRehfD7779z9OhRoGB2VU9Pjy1btgAF+zcfPlDl8OHDQNH/YeJZeOLTLIUQQgghxJP5PuD7yg7hiTz68mcoWMZ2+vRpIiIiaNCgAevXr2f+/PnY29szZcoUZs+erWnr4eFBdHQ0K1eu5Pbt29SvX5/Jkyczffp0oOBl0MuXL2f8+PFkZmZiY2ODr68v8+bNe2xcPXv2ZNeuXcyZM4dBgwZRrVo1vLy8WLhwoeZUxYpQu3ZtYmNjGT9+PIMGDaJu3brMmjWL+Ph4zbvfymrRokU0atSIVatWERERgampKa1atdKc4mlpacmKFSsIDQ3F09OTvLw8EhIS8PLy0umrVatWJCYmMn36dAICAlBVFXd3dw4cOFDk6ZPlNXjwYKKjo1m+fDkAJ06c4NSpU8ydO7fIkyL19fUJDAwkJCSE3377DScnJzw8PDh58iQffvghYWFhXLhwAUNDQ5ydnXn77bd13glXGt27d+fUqVMsWrSIlStXkp6ejqIoODo64uXlxYcffkibNm007RMSEggKCtLq49VXXwXgs88+03qHX2xsLG3btqVRo0ZljqsiKMWd8PK8cHNzU1NTUys7jOdLsFkF9XO9YvoRQgghRInOnDmDs7NzZYchxFPzyy+/0LRpUxITE+ncuXNlh/PU5eTkYGdnx6JFi/jnP//5RH2V9PeDoijHVFXVeQFfRb2aQAghhBBCCPE31rBhQ4KCgliwYEFlh/JMrFmzBhsbGwICAiotBknmhBBCCCGEEBVi3rx5tGvX7qkcNvK8MTIyIjw8XOs9hs+a7JkTQgghhBBCVAhbW1ut/ZD/y0aPHl3ZIcjMnBBCCCGEEEJURZLMCSGEEEIIIUQVJMmcEEIIIYQQQlRBkswJIYQQQgghRBUkyZwQQgghhBBCVEGSzAkhhBBCCCFEFSTJnBBCCCGEEEJUQfKeOSGEEEKISnammXOlju/845knun7Pnj2EhYVx5MgRbt26Rf369RkwYABTp07F3Nxc0y47O5ulS5fi5+eHq6urVh9eXl7k5uaSlJT0RLE8ifz8fFxdXXnjjTeYOHGiTv3w4cNZt24d48ePZ8mSJUX2oaoqX3zxBevXr+e7777jxo0bvPDCC3Tu3JlRo0bh5OSEk5NTibF4enqSmJhIYGAgERERmnJDQ0Pq1atH//79mTlzJrVr19bU9evXj7p167Jy5cpy3L2oiiSZE0IIIYQQ5fbBBx8wffp0+vfvz6effoqFhQXHjh0jNDSUmJgYEhISqFevHlCQzM2ZM4e6devqJHPPg6ioKDIyMnjrrbd06u7cuUN0dDQAGzZsYOHChRgYaP9TOi8vjyFDhrBt2zYCAgJ49913sbCwID09nejoaLp3705WVhbJycla13Xo0IHAwEBGjRqlKatVq5bm19bW1nz11VcA3L17l9TUVGbPns3Zs2fZuXOnpl1wcDDt27dn7NixNGnS5MkfiHjuSTInhBBCCCHKJSEhgRkzZjBu3DjCwsI05Z6engwYMIC2bdvi7+9PQkJCJUb5X3fv3sXIyKjY+kWLFhEQEED16tV16rZt28aNGzfo1asXX3/9NXFxcfTp00erTUhICFu2bGHLli288sorWnWvv/46e/bswdDQEHd3d53+69SpU2Q5QLVq1bTqPD09ycrKIiQkhFu3bmFqagqAi4sLbdq0YenSpaxatar4ByH+Z8ieOSGEEEIIUS4LFy7EwsKCkJAQnTonJyemTp1KYmIihw8fJi0tTbO8cMSIESiKgqIohIeHa10XHx+Pq6sr1atXp0WLFmzfvl2n75MnT+Ln54e5uTkmJiZ06tSJb7/9VqtNYGAgdevWJTk5mY4dO2JiYsLkyZOLvZfDhw/z/fff89prrxVZHxERgbm5OeHh4ZiYmBAZGalVf+/ePRYvXkzv3r11ErlCvr6+RSaK5VGrVi3y8/PJy8vTKh8yZAgbNmzgzp07FTKOeL5JMieEEEIIIcosNzeXAwcO4OPjg7GxcZFt/Pz8ANi/fz92dnZs3boVgPfff5/k5GSSk5Pp3bu3pv0vv/zC2LFjmTBhAlu3bsXOzo6BAwdy7tw5TZvjx4/TsWNHMjMzWbt2LTExMVhaWuLt7c2xY8e0xr9+/TpDhgxh6NChfPPNN8UmagBxcXHUrFmT1q1b69RlZGQQHx/P4MGDsba2pn///nz11VdkZWVp2qSmppKdna2554qWm5tLbm4ut27d4uDBg6xYsYKePXtqLccE8PDw4MaNGzpLOcX/JllmKYQQQgghyuzatWvcuXMHR0fHYtsU1qWnp2NkZISLiwsADRo0KHJJ4dWrVzl48CCNGzcGwNXVFTs7OzZv3sy0adMAmDRpEvXr12f//v1Uq1YNgB49etCiRQvmzZunNZP3119/ERUVRb9+/Uq8n5SUFFq2bImenu5cx+eff05+fj7+/v4ABAQEsHHjRjZt2sTo0aM19wjg4OBQ4lhl9ccff2BoaKhV5u7uzueff67TtnXr1ujp6ZGSkkK3bt0qPBbxfJGZOSGEEEIIUWaqqlZ4n40bN9YkcgA2NjbY2Nhw/vx5oOAQkgMHDvDqq6+ip6enma1SVRVvb28OHjyo1Z+BgYHOvrbiZGRkYG1tXWRdZGQkjRs3pkOHDgB4e3tjb2+vs9TyabGxseHo0aMcPXqU5ORkIiIiuHr1Ki+99JLOckpDQ0PMzMzIyMh4JrGJyiUzc0IIIYQQosysrKwwMTEhLS2t2DaFdYWnWZbEwsJCp8zIyIicnBwAMjMzycvLY968ecybN6/IPvLz8zWzazY2Nujr65dq7JycnCIPRzl69Cj/+c9/mDJlCtnZ2Zryl19+mRUrVnD27FmaNGmiucfff/+9VOOVhaGhIW5ubprv7u7uODs7849//IPw8HCd0zdNTExkz9zfhCRzQgghhBCizAwMDPDw8GDv3r3k5OQUuW+u8Dj9ilruV7t2bfT09Hj77bc1Sx4f9fAySUVRSt23paWl1h64QoXveAsNDSU0NFSnPjIykvnz5+Pm5kbt2rXZuXMnI0eOLPW45dW8eXMATp06pVOXmZmJlZXVU49BVD5ZZimEEEIIIcpl0qRJXLt2TbOf7WG//fYboaGheHh40L59ewDNzFd5Z41MTU3p0qULJ0+exNXVFTc3N51PeTVr1oxff/1Vq+zevXt8+eWXtG/fnoSEBJ1PmzZt+Pzzz1FVlWrVqjFx4kRiY2OJiYkpcoy9e/dy+/btcsf4sMIk7tGloZcuXSInJ4emTZtWyDji+SYzc0IIIYQQoly6d+/O3LlzmTVrFmlpafj7+2Nubs7x48dZsGABZmZmWod0vPDCC1haWvLll1/SqlUrTE1NcXJywtLSstRjLlmyBA8PD3r06ME///lP7OzsuHr1KsePHycvL48FCxaU6148PDz47LPPuHbtmiae2NhYrl27xuLFi/Hy8tK5ZtSoUbz11lskJibStWtX3n//fU6ePMngwYMJDAykb9++WFhYcOHCBWJiYti6dWuRs38luXfvHikpKUDBqZZnz57l3//+NzVr1iQwMFCr7eHDhzX3I/73STInhBBCCFHJnH88U9khlNvMmTNp164dYWFhBAUFcfv2berXr4+/vz/vv/++1j44PT09Pv30U6ZNm4a3tze5ubl89tlnOgnJ47i6unL06FHmzJnDe++9x/Xr17G2tsbV1VVzsmR59OvXD2NjY2JjYwkICAAKlljWrFmTV199tchrhg4dyoQJE4iIiKBr167o6+uzefNmNmzYwLp16wgMDOSvv/7ihRdeoEuXLhw4cAAzM7Myx3blyhXN4Sv6+vrUqVOHjh07Mnv2bBo0aKDVNjY2lrZt29KoUaMyjyOqHuVpnERUkdzc3NTU1NTKDuP5Elz2vwSK7ud6xfQjhBBCiBKdOXMGZ2fnyg5DPEZgYCAXLlwgPj6+skMpl5ycHOzs7Fi0aBH//Oc/KzscUQYl/f2gKMoxVVV11hHLnjkhhBBCCCGA2bNnk5SURFWdSFizZg02NjaamUXxv0+SOSGEEEIIIQAnJyfCw8P5888/KzuUcjEyMiI8PBwDA9lJ9Xchv9NCCCGEEEI8MGTIkMoOodyeZM+gqJpkZk4IIYQQQgghqiBJ5oQQQgghhBCiCip1MqcoSqiiKPsURUlXFOWOoiiZiqKcUBRltqIoRb4cRFGUjoqifP2g7W1FUU4pijJOURT9irsFIYQQQgghhPj7KcvM3HjAFNgLLAM2ALlAMHBKUZR6DzdWFKUfcBDwALYBK4FqQBjw5ZMGLoQQQgghhBB/Z2U5AKWWqqo5jxYqivJvYBrwPjDmQVktYC2QB3ipqpr6oHwmsB8YqCjKEFVVJakTQgghhBBCiHIo9cxcUYncA5sf/Gz8UNlAwBr4sjCRe6iPGQ++vlWGOIUQQgghhBBCPKQiDkDp++DnqYfKuj34GVdE+4PAbaCjoihGFTC+EEIIIYQQQvztlPk9c4qi/AuoAZgBbkBnChK5BQ81a/rg59lHr1dVNVdRlN+A5kAD4ExZYxBCCCGE+F+ycvT+Sh3/7dXdSm70GHv27CEsLIwjR45w69Yt6tevz4ABA5g6dSrm5uaadtnZ2SxduhQ/Pz9cXV21+vDy8iI3N5ekpKQniuVJ5Ofn4+rqyhtvvMHEiRN16ocPH866desYP348S5YsKbIPVVX54osvWL9+Pd999x03btzghRdeoHPnzowaNQonJyecnJxKjMXT05PExERNnxs3buSzzz7jxIkTXL9+ndq1a9O6dWtefvllgoKCMDExAeDmzZvMmTOH1NRUjh8/zs2bN0lISMDLy0ur/xMnTtCpUyd+/PFH6tevX7YHJZ4b5ZmZ+xcwGxhHQSIXB/iqqnrloTZmD35eL6aPwvLaRVUqijJSUZRURVFSr1y5UlQTIYQQQgjxHPjggw/o0aMHxsbGfPrpp+zevZvRo0cTHh5Ou3btSE9P17TNzs5mzpw5HD9+vBIjLl5UVBQZGRm89ZbubqA7d+4QHR0NwIYNG8jNzdVpk5eXx6BBgwgICMDR0ZF169axb98+QkNDycnJoXv37pibm5OcnKz1AQgMDNQqW7VqFQC5ubkMHDgQf39/6taty5o1a9i/fz+rV6/GycmJCRMmsHz5ck0M165dY/369RgYGODj41Psvbq4uODj48PMmTOf6JmJylXmmTlVVW0BFEV5AehIwYzcCUVR+qiqWtr/ZSqF3RUzxifAJwBubm5FthFCCCGEEJUrISGBGTNmMG7cOMLCwjTlnp6eDBgwgLZt2+Lv709CQkIlRvlfd+/excio+F0+ixYtIiAggOrVq+vUbdu2jRs3btCrVy++/vpr4uLi6NOnj1abkJAQtmzZwpYtW3jllVe06l5//XX27NmDoaEh7u7uOv3XqVOnyPL58+ezbds2YmJiGDBggFbdK6+8wvvvv8+PP/6oKXNwcCAzMxOA+Ph4tm7dWuz9jho1in79+hESEoK9vX2x7cTzq9x75lRVvayq6jbAF7AEIh+qLpx5M9O5sECtR9oJIYQQQogqZuHChVhYWBASEqJT5+TkxNSpU0lMTOTw4cOkpaVplheOGDECRVFQFIXw8HCt6+Lj43F1daV69eq0aNGC7du36/R98uRJ/Pz8MDc3x8TEhE6dOvHtt99qtQkMDKRu3bokJyfTsWNHTExMmDx5crH3cvjwYb7//ntee+21IusjIiIwNzcnPDwcExMTIiMjterv3bvH4sWL6d27t04iV8jX17fIRLE4d+/eJSwsjL59++okcoUaNGhAr169NN8VRSmyXXHx1KpVS+f3QFQdT3wAiqqqvwP/AZorimL1oPinBz+bPNpeURQDwImCd9T9+qTjCyGEEEKIZy83N5cDBw7g4+ODsbFxkW38/PwA2L9/P3Z2dppZovfff1+znLB3796a9r/88gtjx45lwoQJbN26FTs7OwYOHMi5c+c0bY4fP07Hjh3JzMxk7dq1xMTEYGlpibe3N8eOHdMa//r16wwZMoShQ4fyzTffFJuoAcTFxVGzZk1at26tU5eRkUF8fDyDBw/G2tqa/v3789VXX5GVlaVpk5qaSnZ2tuaeK0Jqaio3btzQmQGsKAYGBnTo0IG4uKLOLBRVQZmXWRajcF4278HP/cDrQE9g4yNtPYDqwEFVVe9W0PhCCCGEEOIZunbtGnfu3MHR0bHYNoV16enpGBkZ4eLiAhTMJhW1pPDq1ascPHiQxo0L3njl6uqKnZ0dmzdvZtq0aQBMmjSJ+vXrs3//fqpVqwZAjx49aNGiBfPmzdOayfvrr7+IioqiX79+Jd5PSkoKLVu2RE9Pd67j888/Jz8/H39/fwACAgLYuHEjmzZtYvTo0Zp7hIJljhXlwoULADoHlKiqSl5enua7oijo6+uXawwXFxc+/PBD8vPzi7x38Xwr1e+YoijNFEWxLaJc78FLw22AQ6qqFv7niS3AVWCIoihuD7U3BuY/+PrxE0UuhBBCCCEqjapW/LEGjRs31iRyADY2oZRqFQAAIABJREFUNtjY2HD+/Hmg4BCSAwcO8Oqrr6Knp0dubi65/9/evcdZVdX/H399ZLgYIiiMCoaKpWJqmo1hXhDwjjdULPv+VKBEzUwUUhMvkJpCfQNvhJkmqKX49YJfQU0Q0FQI0W+YhXdJxUwugqmggev3xzmMMzDDnLkchj28no/HeeyZvddea50zs9nzZu+91sqVpJQ45JBDePLJJyvVV1JSUvBVrXfffZfS0tIqt91+++3stNNOfPvb3wbgkEMOoVOnTmvdatnQqvuMJ0yYQPPmzctfBx10UJ3bKC0t5dNPPy1/zk7ZUuiVuSOAX0bEk8DrwGJga+AgctMLvAcMXF04pfRhRAwkF+pmRMTdwBLgWHLTFtwLTGioNyFJkqT1q0OHDmy66abMnz+/2jKrt3Xu3LmgOrfccsu11rVs2ZIVK1YAsGTJElatWsWVV17JlVdeWWUdFa8wbbXVVgVfsVqxYkWVg6M8++yz/P3vf+eiiy5i6dKl5etPOOEEbrzxRl555RV23nnn8vf4j3/8o6D2CrG6ztVhdrXDDz+cZ599FsgNYlIfq6c0WL58eb3qUeMoNMxNJTe65P7AnuSmFPiY3DxydwDXp5QqxfmU0sSIOAi4BDgRaAW8BgzOl3eUSkmSpIwqKSmhe/fuTJkyhRUrVlT53Nz//u//AtCrV/3msVutXbt2bLLJJvzoRz8qv+VxTRVvFazNYCDt27ev9AzcauPHjwdg5MiRjBw5cq3tt99+O1dddRVlZWW0a9eOhx56iDPOOKPgdtelrKyMzTffnEmTJjFwYPl1E7bYYgvKynI3v7Vp06bKaRIKtfqKXIcOHWooqQ1RQbdZppReTCn9KKW0V0qpQ0qpJKXUNqW0T0pp+JpBrsJ+T6eUeqeUtkgpbZpS2iOlNDqltKqq8pIkScqOCy64gMWLF5c/z1bRm2++yciRI+nevTvdunUDKL/yVderQK1bt+bAAw9k7ty57L333pSVla31qquuXbvyxhuVx+b77LPPuPvuu+nWrRvTp09f67XXXntxxx13kFKiRYsWDBkyhEmTJnHfffdV2caUKVP45JNPCu5Ty5YtGTRoEA899FCVo3o2hDfffJPOnTuXX6FTtjTUACiSJEnayBx88MFcccUVXH755cyfP5/TTjuNLbbYgueff54RI0bQtm1b7rjjjvLyW2+9Ne3bt+fuu+/m61//Oq1bt6ZLly60b9++4DZHjRpF9+7dOfzww/nBD35Ax44dWbRoEc8//zyrVq1ixIgRdXov3bt357bbbmPx4sXl/Zk0aRKLFy/mV7/6FT169FhrnzPPPJMf/vCHzJgxg549e3LxxRczd+5cvvvd79K/f3+OOeYYttxyS9555x3uu+8+7r///iqv/q3L5ZdfzgsvvFA+cfjRRx/NVlttxbJly5g9ezZz585dazCZRx55hI8//pi//vWvADzxxBMsWrSI1q1bc+SRR1Yq++c//5nu3bvXqk/acBjmJEmSGtmPbmqY2xAbw2WXXcY+++zD6NGjGTBgAJ988gnbbbcdp512GhdffHGl5+A22WQTbrnlFoYOHcohhxzCypUrue222+jfv3/B7e299948++yz/OxnP+Pcc89l2bJllJaWsvfee5ePLFkXxx13HK1atWLSpEn069cPyN1i2aZNG0466aQq9/ne977H4MGDGT9+PD179qRZs2bcc889/P73v+fWW2+lf//+fPTRR2y99dYceOCBPPHEE7RtW900zFUrKSnh/vvv5w9/+AO33XYbAwcO5MMPP6Rdu3bsueeeXH311QwYMKDSPj/84Q8rPbs3fPhwIDfSZsVnHN9++23mzp1b7fOH2vDFhv7oWllZWZozZ05jd2PDMrx2/whUX49ztkuStL7MmzePXXfdtbG7oXXo378/77zzDlOnTm3srqwXI0eOZOzYsbz++ut1ntpADaOmfx8i4rmU0lr3ETuZhCRJkgQMGzaMp556io3hQsKKFSu47rrruOKKKwxyGWaYkyRJkoAuXbowbtw43n///cbuStHNnz+fQYMGceqppzZ2V1QPPjMnSZIk5Z188smN3YX1omvXrnTt2rWxu6F68sqcJEmSJGWQYU6SJEmSMsgwJ0mSJEkZZJiTJEmSpAwyzEmSJElSBhnmJEmSJCmDDHOSJEmSlEHOMydJktTIfvXdoxu1/SETJtVr/5kzZ3Ldddfx1FNP8f7779OqVSt22WUXjjjiCM4++2w6duxYq/pmzJhBz549mT59Oj169KhX3yp69tlnGTFiBE899RRLly6lY8eO9O7dm0suuYRtt922UtmPP/6Ys846iz/+8Y8sXLiQQYMGce2111ZZb0RUuf7//b//x5133tlg/a+NiRMn8sYbbzB48OBK64v12RbqmGOOYYcdduCGG25Ya9tVV13FZZddxvHHH8/9999fbR2PPPIIY8aMYfbs2XzwwQdsueWWdOvWjQEDBnD88cdX+/OoaPvtt2f+/PmV6hw7diyzZ89m8eLFbLbZZuy2224cc8wxnHnmmbRr16687NChQ5kzZw7PPfccS5Ys4bbbbqN///6V6v/nP//JV77yFWbMmMG3vvWtmj+YOjLMSZIkqc5+9atfccEFF9CzZ0+uuuoqdtxxRz766COeeeYZbr75ZubMmcMjjzxSqzr33ntvZs6cyde+9rUG6+cdd9zBgAEDOOCAA7juuuvo1KkT8+bN4xe/+AX33nsvU6dO5etf/3p5+TFjxnDXXXfxu9/9jp133rnGQNq/f3/OPPPMSutKS0sbrP+1NXHiRKZOnbpWmCvGZ1uoJ598kilTpvD6669Xuf32228HYPLkySxevJj27duvVWbIkCGMGjWKvn37cuONN7LNNtvwr3/9i4ceeojvfOc7zJkzh5kzZ1ba5/jjj2fPPfdk+PDh5etatmxZ/vWgQYO4/vrrOeGEExg9ejTbbrstS5cuZdq0aVx99dXMnz+fsWPHlpe/4YYb2GuvvTj66KPL+7ymjh07MnDgQC644AKeeOKJgj+j2jLMSZIkqU6mT5/OBRdcwKBBgxg9enSlbb179+biiy/mf/7nf2pd7+abb86+++7bUN3k5ZdfZuDAgfTp04d77rmHTTbJPWnUvXt3+vbtS7du3ejbty9/+9vfaN68OQDz5s2jU6dOnHbaaQW1se222zZon4uloT/b2vjlL3/JMcccs9ZVUIBnnnmGV199ld69e/Pwww9z1113cc4551Qqc+eddzJq1Cj++7//myFDhlTadtJJJzFo0CC22GIL9txzz0rbWrZsSYcOHap83+PGjeP6669n9OjRnHfeeZW2HXvssVxyySVMnz690vply5axySab8Nprr1Ub5gDOPPNMdtttN2bPnl20q3M+MydJkqQ6GTlyJB06dGDkyJFVbm/duvVat58NGzaMvffem7Zt29KhQwd69erFrFmzKpWZMWMGEcGMGTPK1/Xo0YMDDjiAqVOnsvfee/OlL32J3XffnYkTJ9bYz2uvvZZVq1Zxww03lAe51dq3b8/VV1/Nq6++Wn5rX0Qwbtw43n77bSJirb7U1g477LDW57C6nYpXi4YPH05E8Oqrr3LUUUex2Wabsf3223PFFVfw+eefV9p34cKFnH322XTu3JmWLVvSuXNnTj31VD799FP69+/P+PHjWbBgQXn/d9hhB6DqzzalxOjRo9lll11o0aIFHTt25JxzzuHDDz9cq7+XXnop119/PV26dKFNmzYcdNBB/O1vf6vxM3j33Xd55JFH+K//+q8qt48fP55mzZrx29/+ls6dO1cZkq6++mp23333tYLcat/85jfZbrvtauxLRSNGjGDPPfdcK8itVlpayne+851K69b8HarO1772NfbYYw9uueWWWvWpNgxzkiRJqrWVK1fyxBNPcOihh9KiRYuC91uwYAHnn38+EydOZNy4cWy11VZ0796dF154ocZ9X3/9dQYNGsTgwYO5//776dixI3379uW1115b536PP/44ZWVl1d4qedRRR7HJJpswbdo0IPcM4OGHH84222zDzJkzmTlzJnvvvfc620gpsXLlykqvujr++OPp1asXEydOpE+fPgwbNozx48eXb//ggw/Yb7/9mDBhAoMHD+bhhx/mF7/4Bf/5z3/47LPPuOyyy+jduzelpaXl/X/ggQeqbe+SSy5h8ODBHHrooTz00ENceOGFjBs3jqOOOmqtEHnnnXcyefJkrrvuOm677TbeeustjjvuuBrf75QpU1i1ahUHHHDAWttWrFjBPffcw6GHHkqnTp045ZRTePbZZ5k3b155mXfffZd58+ZxzDHHFPox1mjBggW8/PLLHHXUUQ1W55q6d+/Oo48+WrT6vc1SkiRJtbZ48WJWrFhR5ZWQNf+wLyn54k/OilcpVq1axRFHHMFuu+3GrbfeynXXXbfONhctWsSTTz7JTjvtBOSe/+rYsSP33HMPQ4cOrXa/t99+m29+85vVbm/dujWlpaW8/fbbAOy777506NCBli1bFnxL4tVXX83VV19dad2rr77KV7/61YL2r2jIkCEMGDAAgEMOOYRp06Zx1113la8bPXo0b7zxBnPmzOEb3/hG+X7f+973AGjTpg2lpaW0aNGixv4vWbKEUaNG0a9fP2688UYADj/8cEpLSzn11FOZNGkSxx57bHn55s2bM2nSpPLbUSF3i+Ps2bPZb7/9qm1n1qxZdOrUqcrnCB988EGWLl1afktrv379uOaaaxg/fjwjRowAKP/ZbL/99ut8P7XxzjvvANT4OxwRNGvWrE5tfOMb32DMmDG8++67dOrUqW4dXQevzEmSJKnWUkpVrn/vvfdo3rx5pVfFP4ynTp1Kz549ad++PSUlJTRv3pxXXnmFl19+ucY2d9ppp/IgB7DVVlux1VZb8dZbbxXt/RTq+9//Ps8++2ylV+fOnetU15pXinbfffdK7/Gxxx5jn332qRTk6mrWrFl8+umnnHLKKZXWn3zyyZSUlKw1eMehhx5aKcjtscceADX+DN59991qB4QZP348m2++OX369AFgl112oVu3btx5551rXRlsSNX9zGfNmlXp97c+AXL1e3733XfrXMe6GOYkSZJUax06dKBVq1Zr/RHfoUOH8jAzcODAStuef/55evfuzWabbcatt97KrFmzePbZZ9lzzz1ZsWJFjW1uueWWa61r2bJljft++ctfrjQM/Zo+/vhjFi1aVOfwBbnRC8vKyiq9Ko6YWBtrvs813+PixYv58pe/XOe+VrRkyRKAtW5BLSkpoX379uXb19U3oMafwYoVK6r8PN577z0ee+wxjjrqKD799FOWLl3K0qVLOfHEE1mwYAGPP/44QPnP5h//+Ect3t26ra5zzd/hPfbYo/x3+Oij6zdtyKabbgrA8uXL61VPdQxzkiRJqrWSkhK6d+/OlClT+OyzzyqtXx1m1ryt7L777qOkpIT777+fPn360K1bN8rKyvjggw+K2teDDz6YOXPm8M9//rPK7ZMnT+bzzz+nV69eRWm/VatWlT4jYK2QVBsdOnRgwYIF9e0W8EU4e++99yqtX7lyZbXTA9RF+/btq/w533nnnaxatYq77rqLLbbYovx14YUXApQ/K9ipUyd23XVXHnrooQbpD+RGIN15552ZPHlypfWtW7cu/x2u7/tf/XPu0KFDveqpjmFOkiRJdXLhhReyaNEiLrroooLKf/LJJzRr1qzSpM7Tpk1rkNsk12XQoEFssskm/PjHP17rtr0lS5YwdOhQvvrVr3LCCScUpf3tt9+eF198sdK6SZPqPlH7YYcdxuzZs5k7d261ZVq2bFnQ1aB9992Xli1bcvfdd1daP2HCBFauXMlBBx1U535W1LVrV95+++21nqe8/fbb2X777Zk+ffparyOOOIIHHniAf//730Busu4XX3yRUaNGVdnG//3f/9X6d+miiy5i7ty5NT6vWVdvvvkmLVq0oEuXLkWp3wFQJEmSVCcHH3wwI0aM4Kc//SkvvPACp512Gl26dGHFihW88sor3H333bRu3bo8vB1xxBFce+219O/fnwEDBvDKK69w5ZVXVjnvWEPadddd+c1vfsPpp5/OwQcfzFlnnUXHjh156aWX+MUvfsHSpUuZMmVKpWfBGtLJJ5/M97//fc4//3yOPvpo5s6dy7hx4+pc3/nnn88f/vAHDjnkEC699FL22GMPFi1axIMPPshNN91EmzZt+NrXvsaSJUsYO3YsZWVltGrVqvz5toq23HJLBg8ezDXXXEPr1q3p3bs38+bN49JLL+WAAw5osJEeu3fvzrBhw3jhhRfKRwZ9/vnn+etf/8rw4cPp0aPHWvusWLGCRx99lHvvvZcBAwZwyimn8PzzzzNkyBBmzpzJd77zHbbZZhvef/99Jk+ezB133MGcOXNqNT3B97//fZ577jnOO+88nnzySU466SS23XZbPv74Y1544QWmTp1KmzZtKu3zxBNPsHDhwvKrmXPmzGGzzTYDoG/fvpXK/vnPf2afffahVatWtfm4CmaYkyRJamRDJtT9Kk1ju/DCC9l///257rrrGDp0KAsXLqRVq1bssssufPe73+Wss84qHwnw8MMP5/rrr2fUqFHcd9997L777tx+++1cddVVRe9n//796dq1KyNHjuScc85h2bJldOzYkSOPPJJLLrmkXs/L1aRfv368/fbb3HrrrfzmN7/hwAMP5IEHHqjTSJcA7dq14+mnn+bSSy9lxIgRLF68mK233ppevXqVTxNx+umnM2vWLIYOHcrSpUvZfvvtq31u8Oc//zmlpaXcdNNN/PrXv6Z9+/acdtppXHPNNQXPqVaTAw88kE6dOvHQQw+Vh7nx48cTEVXOwQe5K5CdO3dm/Pjx5SN5jho1ikMOOYQxY8Zw9tlns3TpUrbcckv23Xdf7r///rUmDC/EmDFj6N27N2PHjuXcc8/lgw8+YLPNNmP33Xdn0KBBnHHGGZXKDxs2rNLAMGPGjGHMmDFA5UFVli9fzuOPP77WKKcNKeo7ck+xlZWVpTlz5jR2NzYsw9s2UD3LGqYeSZJUo3nz5rHrrrs2djekRjN8+HB+//vf88orr1S61bapmjBhAqeffjpvv/027dq1W2fZmv59iIjnUkpla673mTlJkiRJRXf++eezdOlS7rvvvsbuynoxcuRILrjgghqDXH0Y5iRJkiQVXdu2bbnjjjvWGtmzKXrvvfc47rjj+MlPflLUdnxmTpIkSdJ6ccQRRzR2F9aLbbbZhmHDhhW9Ha/MSZIkSVIGGeYkSZLWkw194DlJ6199/l0oKMxFRPuIOD0iHoiI1yJieUQsi4inIuIHEbHJGuV3iIi0jtfd1bUlSZLUFLVo0aKgSZwlbVyWL19e5zkOC31m7iRgLPBPYDrwFrA1cAJwC3BkRJyU1o6Vc4GJVdT3Yp16K0mSlFEdOnTgnXfeoUOHDrRp04aSkpKNYnh2SVVLKbF8+XIWLFjA1ltvXac6Cg1zrwDHApNTSp+vXhkRQ4HZwInkgt2a44z+JaU0vE49kyRJakLatm1Ly5YtWbhwIYsXL2blypWN3SVJjax58+ZsvfXWbL755nXav6Awl1KaVs369yLiJuDnQA/WDnOSJEnKa9WqFZ07d27sbkhqIhpiaoL/5JdV/fdSp4g4E2gPLAZmppReaIA2JUmSJGmjVq8wFxElwGn5bx+tosih+VfFfWYA/VJKb9WnbUmSJEnamNV3aoIRwO7AwymlP1ZY/wlwJfBNYIv86yByg6f0AB6PiNbVVRoRZ0TEnIiYs3Dhwnp2UZIkSZKanjqHuYg4FxgCvAScWnFbSun9lNLlKaXnU0pL868ngcOAPwNfBU6vru6U0s0ppbKUUllpaWlduyhJkiRJTVadwlxE/Ai4Dvg70DOltKSQ/VJKK8lNZQDQvS5tS5IkSZLqEOYi4jzgRnJzxfVMKb1XyypW3zdZ7W2WkiRJkqR1q1WYi4iLgNHAX8gFuffr0Oa++eUbddhXkiRJkkQtwlxEXEZuwJPngINTSovWUbZbRLSoYn0v4Pz8t3fWsq+SJEmSpLyCpiaIiH7AFcAq4E/AuRGxZrH5KaVx+a9HArvlpyF4J7/u60Cv/NeXpZSeqXu3JUmSJGnjVug8c13yy2bAedWUeQIYl//6DuB4YB/gSKA58C/gHuDGlNKf6tJZSZIkSVJOQWEupTQcGF5opSmlW4Fb69YlSZIkSVJN6jtpuCRJkiSpERjmJEmSJCmDDHOSJEmSlEGGOUmSJEnKIMOcJEmSJGWQYU6SJEmSMsgwJ0mSJEkZZJiTJEmSpAwyzEmSJElSBpU0dgckSfU35qxp9a7jRzf1aoCeSJKk9cUrc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIyyDAnSZIkSRlkmJMkSZKkDDLMSZIkSVIGGeYkSZIkKYMMc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIyyDAnSZIkSRlkmJMkSZKkDDLMSZIkSVIGGeYkSZIkKYMMc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIyqKAwFxHtI+L0iHggIl6LiOURsSwinoqIH0RElfVExH4R8XBELImITyLihYg4LyKaNezbkCRJkqSNS0mB5U4CxgL/BKYDbwFbAycAtwBHRsRJKaW0eoeIOA64D1gBTACWAMcAo4H983VKkiRJkuqg0DD3CnAsMDml9PnqlRExFJgNnEgu2N2XX7858FtgFdAjpTQnv/4yYBrQNyJOTind3VBvRJIkSZI2JgXdZplSmpZSeqhikMuvfw+4Kf9tjwqb+gKlwN2rg1y+/Arg0vy3P6xrpyVJkiRpY9cQA6D8J79cWWFdr/zy0SrKPwl8AuwXES0boH1JkiRJ2ujUK8xFRAlwWv7bisFtl/zylTX3SSmtBN4kd4vnjvVpX5IkSZI2VvW9MjcC2B14OKX0xwrr2+aXy6rZb/X6dlVtjIgzImJORMxZuHBhPbsoSZIkSU1PncNcRJwLDAFeAk6t7e75ZapqY0rp5pRSWUqprLS0tK5dlCRJkqQmq05hLiJ+BFwH/B3omVJaskaR1Vfe2lK1zdcoJ0mSJEmqhVqHuYg4D7gReJFckHuvimIv55c7V7F/CdCF3IApb9S2fUmSJElSLcNcRFxEbtLvv5ALcu9XU3RafnlEFdu6A18CnkkpfVqb9iVJkiRJOQWHufyE3yOA54CDU0qL1lH8XmARcHJElFWooxVwVf7bsbXvriRJkiQJctMD1Cgi+gFXAKuAPwHnRsSaxeanlMYBpJQ+jIiB5ELdjIi4G1gCHEtu2oJ7gQkN8QYkSZIkaWNUUJgj94wbQDPgvGrKPAGMW/1NSmliRBwEXAKcCLQCXgMGA9enlKocyVKSJEmSVLOCwlxKaTgwvLaVp5SeBnrXdj9JkiRJ0rrVd9JwSZIkSVIjMMxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlkGFOkiRJkjLIMCdJkiRJGWSYkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlkGFOkiRJkjLIMCdJkiRJGWSYkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlkGFOkiRJkjLIMCdJkiRJGWSYkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlUEFhLiL6RsQNEfGniPgwIlJE3FlN2R3y26t73d2wb0GSJEmSNj4lBZa7FNgT+Ah4B+hawD5zgYlVrH+xwDYlSZIkSdUoNMydTy7EvQYcBEwvYJ+/pJSG17FfkiRJkqR1KCjMpZTKw1tEFK83kiRJkqSCFHplri46RcSZQHtgMTAzpfRCEduTJEmSpI1GMcPcoflXuYiYAfRLKb1VxHYlSZIkqckrxtQEnwBXAt8Etsi/Vj9n1wN4PCJar6uCiDgjIuZExJyFCxcWoYuSJEmSlG0NHuZSSu+nlC5PKT2fUlqafz0JHAb8GfgqcHoNddycUipLKZWVlpY2dBclSZIkKfPW26ThKaWVwC35b7uvr3YlSZIkqSlab2Eub/U9k+u8zVKSJEmStG7rO8ztm1++sZ7blSRJkqQmpcHDXER0i4gWVazvRW7ycYA7G7pdSZIkSdqYFDQ1QUT0Afrkv90mv/x2RIzLf70opfST/Ncjgd3y0xC8k1/3daBX/uvLUkrP1KfTkiRJkrSxK3Seub2Afmus2zH/AvgHsDrM3QEcD+wDHAk0B/4F3APcmFL6U306LEmSJEkqMMyllIYDwwsseytwa927JEmSJEmqyfoeAEWSJEmS1AAMc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIyyDAnSZIkSRlkmJMkSZKkDDLMSZIkSVIGGeYkSZIkKYMMc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIyyDAnSZIkSRlkmJMkSZKkDDLMSZIkSVIGGeYkSZIkKYMMc5IkSZKUQSWN3QFJkiRJjWde110bpJ5dX5rXIPWocF6ZkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlkGFOkiRJkjLIMCdJkiRJGWSYkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBBYW5iOgbETdExJ8i4sOISBFxZw377BcRD0fEkoj4JCJeiIjzIqJZw3RdkiRJkjZeJQWWuxTYE/gIeAfouq7CEXEccB+wApgALAGOAUYD+wMn1bG/kiRJkiQKv83yfGBnYHPgh+sqGBGbA78FVgE9Uko/SCldAOwFzAT6RsTJde+yJEmSJKmgMJdSmp5SejWllAoo3hcoBe5OKc2pUMcKclf4oIbMvoS1AAARpUlEQVRAKEmSJElat2IMgNIrv3y0im1PAp8A+0VEyyK0LUmSJEkbhWKEuV3yy1fW3JBSWgm8Se5ZvR2rqyAizoiIORExZ+HChUXooiRJkiRlWzHCXNv8clk121evb1ddBSmlm1NKZSmlstLS0gbtnCRJkiQ1BY0xz1zkl4U8fydJkiRJqkIxwtzqK29tq9m++RrlJEmSJEm1VIww93J+ufOaGyKiBOgCrATeKELbkiRJkrRRKEaYm5ZfHlHFtu7Al4BnUkqfFqFtSZIkSdooFCPM3QssAk6OiLLVKyOiFXBV/tuxRWhXkiRJkjYaJYUUiog+QJ/8t9vkl9+OiHH5rxellH4CkFL6MCIGkgt1MyLibmAJcCy5aQvuBSY0TPclSZIkaeNUUJgD9gL6rbFuR76YK+4fwE9Wb0gpTYyIg4BLgBOBVsBrwGDg+pSSI1lKkiRJUj0UFOZSSsOB4bWpOKX0NNC79l2SJEmSJNWkMeaZkyRJkiTVk2FOkiRJkjLIMCdJkiRJGWSYkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlkGFOkiRJkjLIMCdJkiRJGWSYkyRJkqQMMsxJkiRJUgYZ5iRJkiQpgwxzkiRJkpRBhjlJkiRJyiDDnCRJkiRlkGFOkiRJkjKopLE7IEmS1q8dfjq53nXMH3FUA/REklQfXpmTJEmSpAwyzEmSJElSBhnmJEmSJCmDDHOSJEmSlEGGOUmSJEnKIMOcJEmSJGWQYU6SJEmSMsgwJ0mSJEkZZJiTJEmSpAwyzEmSJElSBhnmJEmSJCmDihbmImJ+RKRqXu8Vq11JkiRJ2hiUFLn+ZcC1Vaz/qMjtSpIkSVKTVuwwtzSlNLzIbUiSJEnSRsdn5iRJkiQpg4p9Za5lRJwCbAd8DLwAPJlSWlXkdiVJkiSpSSt2mNsGuGONdW9GxICU0hNFbluSJEmSmqxihrnbgD8BfwP+DewInAOcATwSEd9OKc2taseIOCNfju22266IXZQkSYIxZ01rkHp+dFOvBqlHG4HhbRuonmUNU48yqWjPzKWUfpZSmpZS+ldK6ZOU0osppbOAUcCmwPB17HtzSqkspVRWWlparC5KkiRJUmY1xgAoN+WX3RuhbUmSJElqEhojzL2fX7ZuhLYlSZIkqUlojDD37fzyjUZoW5IkSZKahKKEuYjYLSK2rGL99sCN+W/vLEbbkiRJkrQxKNZolicBP42I6cCb5Eaz/ApwFNAKeBj47yK1LUmSJElNXrHC3HRgF+Ab5G6rbA0sBZ4iN+/cHSmlVKS2JUmSJKnJK0qYy08I7qTgkiRJklQkxZw0XBuBeV13bZB6dn1pXoPUI0mSpMYx5qxp9a7jRzf1aoCebDwaYzRLSZIkSVI9GeYkSZIkKYMMc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIyyDAnSZIkSRlkmJMkSZKkDHLScG0QnGRSktQU/Oq7R9e7jiETJjVATyRtDLwyJ0mSJEkZZJiTJEmSpAwyzEmSJElSBhnmJEmSJCmDDHOSJEmSlEGGOUmSJEnKIMOcJEmSJGWQYU6SJEmSMshJwyVJUqbN67pr/SvpMab+dUjSeuaVOUmSJEnKIMOcJEmSJGWQYU6SJEmSMsgwJ0mSJEkZZJiTJEmSpAwyzEmSJElSBhnmJEmSJCmDDHOSJEmSlEFOGr6e7fDTyfWuY36rBugIsMf4Pepdxz0N0A9JkqQs2ZD+ntPGzStzkiRJkpRBhjlJkiRJyqCihrmI+HJE/C4i3o2ITyNifkRcGxFbFLNdSZIkSWrqivbMXER8BXgG2Ap4EHgJ+BYwCDgiIvZPKS0uVvuSJEmS1JQV88rcr8kFuXNTSn1SSj9NKfUCRgO7AD8vYtuSJEmS1KQVJcxFxI7AYcB8YMwam4cBHwOnRkTrYrQvSZIkSU1dsa7M9covH0spfV5xQ0rp38DTwJeAfYvUviRJkiQ1acUKc7vkl69Us/3V/HLnIrUvSZIkSU1apJQavtKIm4GBwMCU0i1VbP85MBQYmlK6portZwBn5L/dBXi5wTuphtIBWNTYnZAyzuNIqh+PIal+PIY2fNunlErXXFm00SxrEPlllUkypXQzcPP6647qKiLmpJTKGrsfUpZ5HEn14zEk1Y/HUHYV6zbLZfll22q2b75GOUmSJElSLRQrzK2+LbK6Z+J2yi+re6ZOkiRJkrQOxQpz0/PLwyKiUhsR0QbYH1gOzCpS+1p/vB1Wqj+PI6l+PIak+vEYyqiiDIACEBF/JDfX3LkppRsqrB8FnA/8JqV0VlEalyRJkqQmrphh7ivAM8BWwIPAPKAb0JPc7ZX7pZQWF6VxSZIkSWriihbmACKiM3AFcATQHvgnMBH4WUppSdEaliRJkqQmrqhhTpIkSZJUHMUaAEWSJEmSVESGOUmSJEnKoJLG7oCyIyI2BfYlN39gOyCRm/j9FWBWSml5I3ZPkrQR8Fwk1Y/HUNPiM3OqUURsAfwcOBX4UjXFlgPjgUtTSh+sr75JWeNJVKobz0VS/XgMNU2GOa1TRLQjN8VEV+Bj4GngVXJ/fAawObATuYngWwMvkZt2YmmjdFjaQHkSlerOc5FUPx5DTZe3Waomw8gd+KOBYSmlj6oqFBGbkZuG4jzgcmDweuuhtIHLn0Sf5ouT6BSqP4n+EOgZEZ5EpS94LpLqx2OoifLKnNYpIt4EXk8pHVJg+WlAl5RSl+L2TMqOiBgNDKJ2J9FrU0qeRCU8F0n15THUdDmapWrSEZhdi/Kz8vtI+kIfYFpKaUh1QQ4gpfRRPsDNAI5fX52TMsBzkVQ/HkNNlGFONVkM7FKL8rvm95H0BU+iUv14LpLqx2OoiTLMqSZ/BPpExNk1FYyIc4BjgUeL3ispWzyJSvXjuUiqH4+hJspn5rROEbEt8DzQAZgPPEZuCPVl+SJtyQ2xfhiwA/A+UJZSWrC++yptqCLid0A/4McppV/XUPYc4DpgXErpB+ujf9KGznORVD8eQ02XYU41iogdgbHAoflVa/7SRH75GHB2SumN9dU3KQs8iUr157lIqh+PoabJMKeCRUQXoBe528Xa5lcvA14GpnvQS9XzJCo1DM9FUv14DDUthjlJWo88iUqSpIZimJMkSZKkDHI0S0mSJEnKIMOcGkxE7BgRb0TE643dF0nSxslzkVQ/HkPZUtLYHVCT0pzcSHzeuyvVUX6glKlASil9pbH7I2WQ5yKpfjyGMsQwp4b0OtClsTshZZwnUal+PBdJ9eMxlCEOgCJJG5CIKAG2BUgp/aORuyNJkjZghjkVLCI2BfYlN7lxO3JXDpaRm/x4VkppeSN2T5IkSXUUEVsCq1JKyxq7LyqcYU41iogtgJ8DpwJfqqbYcmA8cGlK6YP11TdJ0sYjIo4HegArgUdTSlOqKdcP6JdS6rUeuydt0CJiW+BiYCdgLjAypbQ4IvYCbgd2yxd9GjgjpfRS4/RUtWGY0zpFRDvgGaAr8DG5A/xVclfkAtic3D8K+wOtgZeA/VJKSxulw1ITEBG/BE5wABQpJyICmACcSO7cA7m7QyYDp615zomIYcDlKaVm67Wj0gYqf9VtLvnb+PP+AhyeX7YH/g50ArYC3gV29++5DZ8DoKgmw8gFudHAsJTSR1UViojNgCuA84DLgcHrrYdS09OB3CAoknIGAH2Bt4GbgP8A/YCjgacioldK6f1G7J+0oTuHXJC7GrgHOJbc323jgH8D304pvQUQET8ndwXvx8CVjdFZFc4rc1qniHgTeD2ldEiB5acBXVJKjoIk1VFE3EbuaoNXFSQgIv5E7hawrqtDW0Q0A0aS+8/DF4FeKaVF+W1emZMqiIjngf+klLpVWPckuTurTkgpPVhhfZC7C2txxfLaMHllTjXpCNxVi/KzgP2K1BcpkyLi9lru4jEkVbYHcG/Fq28ppVXATyLiLeBaYGpE9PS5balK2wN3rrFuDrkw90zFlSmlFBFPACesp76pHgxzqsliYJdalN81v4+kL5xC7vmeqKlgBd42IX2hBfCvqjaklK6PiFXADcCUiCjoThJpI7MpubEPKloGkFJaWEX5f5EbC0EbOMOcavJHoF9EnJ1S+vW6CkbEOeTuwR63PjomZci/gXeAswss/1PgsOJ1R8qcBcB21W1MKY2JiObAKHLnrafXV8ekjFhEbmCTij4GqnvWtD3g4CcZ4DNzWqf8MLbPkxuQYT7wGLl55VbPQdKW3Lxzh5EbsOF9oCyltGB991XaUOWfS9gzpdS2wPI+MydVEBH3A99KKX25hnIXAdeQm7qgmceQlBMRU4EvpZQKuo0/X37LlNLexe2Z6ssrc1qnlNKCiPg2MBY4FDiTtW//Wn3r2GPA2QY5aS1/AfaPiK+klF5v7M5IGfQw0CcijkopTa6uUEppZES0AH6GtypLFT0HnBcRLVJKn62rYERsDXQHfrNeeqZ6McypRimlN4DDI6IL0IvcM3SrrzAsA14GpufLSVrbE8CBwJeBQsLcRHJXwiXl3A80Y+1nftaSUroyPyjKDsXulJQVKaWLgIsKLN4OuACYWrweqaF4m6UkSZIkZdAmjd0BSZIkSVLtGeYkSZIkKYMMc5KkzIuIHSIiRcS4xu6LJEnri2FOkqQ6iIjh+QDZo7H7IknaOBnmJEmSJCmDDHOSJEmSlEGGOUnSBi8ivhUREyJiQUR8GhH/jIjHIuI7New3IyKqnIMnIvrnb5Psv8b6r0fEXRExP9/Wwoh4PiKujYjm+TLzgWH5Xabn60lrthURX4qIiyPiLxHxcUR8FBEzI+J7VfSnR76O4fn3OzkiluTX7VDoZyVJ2ng4abgkaYMWEQOBscAq4H+BV4GtgDLgbOCeBmzr68CfgZRv601gc+Cr+bYuBf4DXAv0AQ4CxlPFJO8R0Q6YBnwDeB74Hbn/RD0c+ENE7JZSurSKbnwbuBh4Kr9PB+CzhnqPkqSmwzAnSdpgRcTXgF8DHwIHppT+tsb2Lzdwk/2AVkCflNKDa7S1BfAJQErp2nxYOwgYl1KaUUVd15ILchellH5RoZ5WwERgaETcm1L6yxr7HQaclVL6TQO9J0lSE+VtlpKkDdkPyf3H45VrBjmAlNI7RWp3eRVtfZBS+ryQnSOiPXAKMKdikMvXswK4CAjgv6rY/S8GOUlSIbwyJ0nakO2bXz6yntqbAAwCJkbEvcBU4OmU0uu1rGcfoBmQImJ4Fdub55e7VrFtdi3bkiRtpAxzkqQNWbv8csH6aCylNDsiDgQuAfoCpwJExMvAz1JKdxVYVfv8cp/8qzqbVbHuvQLbkCRt5LzNUpK0IVuaX25bx/0/B4iIqv7zsl0V60gpzUwpHQ1sAewPXAlsTW7QkkMKbHdZfjk6pRTrePWsqgsFtiFJ2sgZ5iRJG7JZ+eWRddz/g/yycxXbyta1Y0rp05TSMymly4Fz86uPq1BkVX7ZrIrdZ5MLkgfWoq+SJNWKYU6StCEbC6wELsuPbFlJAaNZrn7+bOAa+x0MVDXX24ER0baKerbOLz+psG5xfrndmoVTSu8DvwfKIuKyqq4MRsRXIqJLDf2XJKlaPjMnSdpgpZT+HhFnAzcB/xcRD5KbZ649uStr/waqulVxtduAC4CLI2JP4O/AzuSu9D0AnLhG+SHAYRExA3gD+AjYLV/+A+DmCmWnk7v6dk1E7J7fTkrpqvz2c4CdgCuAUyPiKeBfQCdyA5/sQy5Qvln4JyJJ0hcMc5KkDVpK6bcR8SLwE6AHucm6FwEvALfUsO/7EXEQ8EugO7l54eYAhwJdWDvM/ZpcKOtG7nm5EuCd/PpfpZT+UaHueRHRL9+vs8nNTwdwVX77h/m2zyA3BcGJ+TL/IhdIzwem1O7TkCTpC5GSz1lLkiRJUtb4zJwkSZIkZZBhTpIkSZIyyDAnSZIkSRlkmJMkSZKkDDLMSZIkSVIGGeYkSZIkKYMMc5IkSZKUQYY5SZIkScogw5wkSZIkZZBhTpIkSZIy6P8DWl1vLM4EQ6wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "#X = df_actg1.loc[:, cols].values # df_actg1.loc[:, cols].values # \n",
    "#X_trans = pca.fit_transform(X)\n",
    "plt.rcParams.update({'font.size': 20})\n",
    "\n",
    "\n",
    "\n",
    "d = {'cat': df_all['Category'], 'cluster': df_hrss['cluster']}\n",
    "df_bar = pd.DataFrame(data=d)\n",
    "#df_bar = df_bar.sample(frac=1)\n",
    "fig, ax = plt.subplots(figsize=(15,7))\n",
    "df_bar.groupby(['cluster', 'cat']).size().unstack().plot(ax=ax, kind = 'bar') \n",
    "plt.legend(loc=1, prop={'size': 16})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rank phenotypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "explore_df =  pd.concat([df_pheno, df_hrss['cluster'].astype(str)], axis=1) #  pd.Series([1 for i in range(len(df_pheno))], name='cluster')\n",
    "l_ignore = ['Category', 'Id', 'Protein', 'Origin'] # , 'cluster'\n",
    "columns = list(explore_df.loc[:, ~explore_df.columns.isin(l_ignore)])\n",
    "\n",
    "l_ignore.append('cluster')\n",
    "columns_raw = list(explore_df.loc[:, ~explore_df.columns.isin(l_ignore)])\n",
    "#explore_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Rank by occurrence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "import obonet\n",
    "\n",
    "url = 'phenopy_mod/.phenopy/data/hp.obo'\n",
    "graph = obonet.read_obo(url)\n",
    "\n",
    "id_to_name = {id_: data.get('name') for id_, data in graph.nodes(data=True)}\n",
    "name_to_id = {data['name']: id_ for id_, data in graph.nodes(data=True) if 'name' in data}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "if dataset == 'IARS':\n",
    "    lbls = {col : id_to_name[col] for col in columns_raw }\n",
    "else :\n",
    "    lbls = columns_raw\n",
    "#lbls"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create HPO2Gene file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "#df_hpoa[['HPO_ID', 'DiseaseName']].to_csv(\"hpo/disease_2_pheno.txt\", sep='\\t', header=False, index=False)\n",
    "df_hpo[['#Format: HPO-id', 'entrez-gene-symbol']].to_csv(\"hpo/pheno_2_gene.txt\", sep='\\t', header=False, index=False)\n",
    "#annotated_table.to_csv(\"results/%s/2_phenotypes/Table_%s_%s_%s.csv\" % (title, str(index), pheno, title), sep='|', index=False)\n",
    "#explore_df = explore_df.loc[:, (explore_df==0).mean() < .8]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Import table that links phenotypes to genes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Load HPO to Genes file\n",
    "df_hpo = pd.read_table(r'hpo/util/annotation/phenotype_to_genes.txt', sep='\\t|<tab>', engine='python') # phenotype_to_genes.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate weight (nr of associated genes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "persistent embryonic sructures not found\n",
      "polyarthralgia not found\n"
     ]
    }
   ],
   "source": [
    "import columns_linked_to_hpo\n",
    "#imp.reload(func)\n",
    "\n",
    "## Calculate Weights\n",
    "l_cols = df_pheno.columns\n",
    "\n",
    "if dataset == 'IARS':\n",
    "    #columns_raw = columns_raw[:-1]\n",
    "    weight_list = func.getNumberOfGenes(df_pheno[columns_raw].columns, df_hpo=df_hpo)\n",
    "else : \n",
    "    weight_list = func.getNumberOfGenesTranslation(df_pheno[columns_raw].columns, col2hpo=columns_linked_to_hpo.column_to_hpo, df_hpo=df_hpo)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### How many patients \n",
    "- to assess the fraction of patients"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "#l_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from numpy.linalg import norm \n",
    "imp.reload(func)\n",
    "\n",
    "name='IARS'\n",
    "\n",
    "n = len(explore_df)\n",
    "\n",
    "\n",
    "df = explore_df[columns].copy()\n",
    "    \n",
    "k = func.defineOptimalNrOfPatients(df, 10)\n",
    "\n",
    "print(k)\n",
    "#plt.xlim((0,40))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.0, 9.0)"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "l_y = func.plotPhenotypesPatients(explore_df[columns], name=dataset)\n",
    "plt.xlim(0,9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Category</th>\n",
       "      <th>Id</th>\n",
       "      <th>Protein</th>\n",
       "      <th>High frequency hearing loss</th>\n",
       "      <th>Short stature</th>\n",
       "      <th>microcephaly</th>\n",
       "      <th>hearing loss</th>\n",
       "      <th>ID</th>\n",
       "      <th>Developmental delay (Global)</th>\n",
       "      <th>Motor Delay</th>\n",
       "      <th>...</th>\n",
       "      <th>Polyarthralgia</th>\n",
       "      <th>Hepatomegaly</th>\n",
       "      <th>Hypothyroidism</th>\n",
       "      <th>Leukocytosis</th>\n",
       "      <th>thrombocytopenia</th>\n",
       "      <th>Asthma</th>\n",
       "      <th>Strabismus</th>\n",
       "      <th>Growth hormone deficiency</th>\n",
       "      <th>Origin</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Patient B</td>\n",
       "      <td>p.Pro32Ser</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family 6</td>\n",
       "      <td>p.Pro264Leu</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Chinese family 1</td>\n",
       "      <td>p.Ile122Val</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Chinese family 2</td>\n",
       "      <td>p.Lys213Arg</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family #1</td>\n",
       "      <td>p.Ile34Met</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family #2</td>\n",
       "      <td>p.Arg37His</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family #4</td>\n",
       "      <td>p.Met82Ile</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family #8</td>\n",
       "      <td>p.Ile165Val</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family #11</td>\n",
       "      <td>p.His275Tyr</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family #14/#15</td>\n",
       "      <td>p.Pro332Ser</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Japanese family 1</td>\n",
       "      <td>p.Gly48Arg</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Japanese family 2/3</td>\n",
       "      <td>p.Lys118Met</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Japanese family 4</td>\n",
       "      <td>p.Glu241Lys</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Japanese family 5</td>\n",
       "      <td>p.Leu299Val</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Scandinavian family</td>\n",
       "      <td>p.Val370Ala</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Dutch family 1</td>\n",
       "      <td>p.Thr278Ile</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Dutch family 2</td>\n",
       "      <td>p.Asp51Asn</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family S840</td>\n",
       "      <td>p.Lys118Asn</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family S582</td>\n",
       "      <td>p.Glu241Lys</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family KNUF21</td>\n",
       "      <td>p.Asp187His</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family MSUDF1</td>\n",
       "      <td>p.Thr89Ile</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family 1250</td>\n",
       "      <td>p.Lys118Met</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>Other (ACTG1)</td>\n",
       "      <td>Family 1320</td>\n",
       "      <td>p.Pro332Ala</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>ACTG1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>23 rows × 98 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         Category                   Id      Protein  \\\n",
       "11  Other (ACTG1)            Patient B   p.Pro32Ser   \n",
       "12  Other (ACTG1)             Family 6  p.Pro264Leu   \n",
       "13  Other (ACTG1)     Chinese family 1  p.Ile122Val   \n",
       "14  Other (ACTG1)     Chinese family 2  p.Lys213Arg   \n",
       "15  Other (ACTG1)            Family #1   p.Ile34Met   \n",
       "16  Other (ACTG1)            Family #2   p.Arg37His   \n",
       "17  Other (ACTG1)            Family #4   p.Met82Ile   \n",
       "18  Other (ACTG1)            Family #8  p.Ile165Val   \n",
       "19  Other (ACTG1)           Family #11  p.His275Tyr   \n",
       "20  Other (ACTG1)       Family #14/#15  p.Pro332Ser   \n",
       "21  Other (ACTG1)    Japanese family 1   p.Gly48Arg   \n",
       "22  Other (ACTG1)  Japanese family 2/3  p.Lys118Met   \n",
       "23  Other (ACTG1)    Japanese family 4  p.Glu241Lys   \n",
       "24  Other (ACTG1)    Japanese family 5  p.Leu299Val   \n",
       "25  Other (ACTG1)  Scandinavian family  p.Val370Ala   \n",
       "31  Other (ACTG1)       Dutch family 1  p.Thr278Ile   \n",
       "32  Other (ACTG1)       Dutch family 2   p.Asp51Asn   \n",
       "33  Other (ACTG1)          Family S840  p.Lys118Asn   \n",
       "34  Other (ACTG1)          Family S582  p.Glu241Lys   \n",
       "35  Other (ACTG1)        Family KNUF21  p.Asp187His   \n",
       "36  Other (ACTG1)        Family MSUDF1   p.Thr89Ile   \n",
       "37  Other (ACTG1)          Family 1250  p.Lys118Met   \n",
       "38  Other (ACTG1)          Family 1320  p.Pro332Ala   \n",
       "\n",
       "    High frequency hearing loss  Short stature   microcephaly   hearing loss  \\\n",
       "11                            1              0              0              1   \n",
       "12                            1              0              0              1   \n",
       "13                            1              0              0              1   \n",
       "14                            1              0              0              1   \n",
       "15                            1              0              0              1   \n",
       "16                            1              0              0              1   \n",
       "17                            1              0              0              1   \n",
       "18                            1              0              0              1   \n",
       "19                            1              0              0              1   \n",
       "20                            1              0              0              1   \n",
       "21                            1              0              0              1   \n",
       "22                            1              0              0              1   \n",
       "23                            1              0              0              1   \n",
       "24                            1              0              0              1   \n",
       "25                            1              0              0              1   \n",
       "31                            1              0              0              1   \n",
       "32                            1              0              0              1   \n",
       "33                            1              0              0              1   \n",
       "34                            1              0              0              1   \n",
       "35                            1              0              0              1   \n",
       "36                            1              0              0              1   \n",
       "37                            1              0              0              1   \n",
       "38                            1              0              0              1   \n",
       "\n",
       "      ID  Developmental delay (Global)   Motor Delay  ...  Polyarthralgia  \\\n",
       "11     0                             0             0  ...             0.0   \n",
       "12     0                             0             0  ...             0.0   \n",
       "13     0                             0             0  ...             0.0   \n",
       "14     0                             0             0  ...             0.0   \n",
       "15     0                             0             0  ...             0.0   \n",
       "16     0                             0             0  ...             0.0   \n",
       "17     0                             0             0  ...             0.0   \n",
       "18     0                             0             0  ...             0.0   \n",
       "19     0                             0             0  ...             0.0   \n",
       "20     0                             0             0  ...             0.0   \n",
       "21     0                             0             0  ...             0.0   \n",
       "22     0                             0             0  ...             0.0   \n",
       "23     1                             1             0  ...             0.0   \n",
       "24     0                             0             0  ...             0.0   \n",
       "25     0                             0             0  ...             0.0   \n",
       "31     0                             0             0  ...             0.0   \n",
       "32     0                             0             0  ...             0.0   \n",
       "33     0                             0             0  ...             0.0   \n",
       "34     0                             0             0  ...             0.0   \n",
       "35     0                             0             0  ...             0.0   \n",
       "36     0                             0             0  ...             0.0   \n",
       "37     0                             0             0  ...             0.0   \n",
       "38     0                             0             0  ...             0.0   \n",
       "\n",
       "    Hepatomegaly  Hypothyroidism  Leukocytosis  thrombocytopenia  Asthma  \\\n",
       "11           0.0               0             0                 0       0   \n",
       "12           0.0               0             0                 0       0   \n",
       "13           0.0               0             0                 0       0   \n",
       "14           0.0               0             0                 0       0   \n",
       "15           0.0               0             0                 0       0   \n",
       "16           0.0               0             0                 0       0   \n",
       "17           0.0               0             0                 0       0   \n",
       "18           0.0               0             0                 0       0   \n",
       "19           0.0               0             0                 0       0   \n",
       "20           0.0               0             0                 0       0   \n",
       "21           0.0               0             0                 0       0   \n",
       "22           0.0               0             0                 0       0   \n",
       "23           0.0               0             0                 0       0   \n",
       "24           0.0               0             0                 0       0   \n",
       "25           0.0               0             0                 0       0   \n",
       "31           0.0               0             0                 0       0   \n",
       "32           0.0               0             0                 0       0   \n",
       "33           0.0               0             0                 0       0   \n",
       "34           0.0               0             0                 0       0   \n",
       "35           0.0               0             0                 0       0   \n",
       "36           0.0               0             0                 0       0   \n",
       "37           0.0               0             0                 0       0   \n",
       "38           0.0               0             0                 0       0   \n",
       "\n",
       "    Strabismus  Growth hormone deficiency  Origin  cluster  \n",
       "11           0                          0   ACTG1        1  \n",
       "12           0                          0   ACTG1        1  \n",
       "13           0                          0   ACTG1        1  \n",
       "14           0                          0   ACTG1        1  \n",
       "15           0                          0   ACTG1        1  \n",
       "16           0                          0   ACTG1        1  \n",
       "17           0                          0   ACTG1        1  \n",
       "18           0                          0   ACTG1        1  \n",
       "19           0                          0   ACTG1        1  \n",
       "20           0                          0   ACTG1        1  \n",
       "21           0                          0   ACTG1        1  \n",
       "22           0                          0   ACTG1        1  \n",
       "23           0                          0   ACTG1        1  \n",
       "24           0                          0   ACTG1        1  \n",
       "25           0                          0   ACTG1        1  \n",
       "31           0                          0   ACTG1        1  \n",
       "32           0                          0   ACTG1        1  \n",
       "33           0                          0   ACTG1        1  \n",
       "34           0                          0   ACTG1        1  \n",
       "35           0                          0   ACTG1        1  \n",
       "36           0                          0   ACTG1        1  \n",
       "37           0                          0   ACTG1        1  \n",
       "38           0                          0   ACTG1        1  \n",
       "\n",
       "[23 rows x 98 columns]"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "explore_df[explore_df['cluster']=='1']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### calculate Phenotypic Specificity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x648 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "imp.reload(func)\n",
    "\n",
    "cluster = 3\n",
    "N = len(explore_df[explore_df['cluster']==str(cluster)])\n",
    "\n",
    "prob = func.PhenotypicSpecifictyOccurrence(explore_df[columns], weights=weight_list, cluster=str(cluster), topN=10, pat_frac=1, rename=False, divide_occurrence=True) # labels=lbls, \n",
    "pheno_ACTG = list(prob.index) # topN=10, \n",
    "\n",
    "prob.plot(kind='barh',figsize=(9,9), title='Most specific phenotypic features in Cluster %s (n=%s)' % (cluster, str(N)))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [],
   "source": [
    "#prob.to_csv(\"hpo/disease_2_pheno.txt\", sep='\\t', header=False, index=False)\n",
    "#pheno_ACTG\n",
    "\n",
    "def translateToHPO(l_cols, col2hpo):\n",
    "    \"\"\"\n",
    "    l_cols = column names (written phenotypes)\n",
    "    \"\"\"\n",
    "    hpo_list = []\n",
    "    for val in l_cols:\n",
    "        col = val.lower().strip()\n",
    "        #print(col)\n",
    "        if col in col2hpo.keys():\n",
    "            if type(col2hpo[col]) == list:\n",
    "                hpo_list.extend([hpo for hpo in col2hpo[col]])\n",
    "            else :\n",
    "                hpo_list.append(col2hpo[col])\n",
    "        #print(hpo)\n",
    "    return hpo_list\n",
    "\n",
    "col2hpo = columns_linked_to_hpo.column_to_hpo\n",
    "\n",
    "hpo_list = translateToHPO(pheno_ACTG, col2hpo) # pheno_ACTG (filtered) or df_pheno.columns\n",
    "#hpo_list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explore pathophysiology"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### load HPOA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Nr of HPO codes: 199197\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\levrex\\Anaconda3\\envs\\my_env\\lib\\site-packages\\IPython\\core\\interactiveshell.py:3063: DtypeWarning: Columns (6,8,9) have mixed types.Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>#DatabaseID</th>\n",
       "      <th>DiseaseName</th>\n",
       "      <th>Qualifier</th>\n",
       "      <th>HPO_ID</th>\n",
       "      <th>Reference</th>\n",
       "      <th>Evidence</th>\n",
       "      <th>Onset</th>\n",
       "      <th>Frequency</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Modifier</th>\n",
       "      <th>Aspect</th>\n",
       "      <th>Biocuration</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>OMIM:210100</td>\n",
       "      <td>BETA-AMINOISOBUTYRIC ACID, URINARY EXCRETION OF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HP:0000007</td>\n",
       "      <td>OMIM:210100</td>\n",
       "      <td>IEA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>I</td>\n",
       "      <td>HPO:iea[2009-02-17]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>OMIM:210100</td>\n",
       "      <td>BETA-AMINOISOBUTYRIC ACID, URINARY EXCRETION OF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HP:0003355</td>\n",
       "      <td>OMIM:210100</td>\n",
       "      <td>IEA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>P</td>\n",
       "      <td>HPO:skoehler[2009-02-17]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>OMIM:163600</td>\n",
       "      <td>NIPPLES INVERTED</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HP:0000006</td>\n",
       "      <td>OMIM:163600</td>\n",
       "      <td>IEA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>I</td>\n",
       "      <td>HPO:iea[2009-02-17]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>OMIM:163600</td>\n",
       "      <td>NIPPLES INVERTED</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HP:0003186</td>\n",
       "      <td>OMIM:163600</td>\n",
       "      <td>IEA</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>P</td>\n",
       "      <td>HPO:iea[2009-02-17]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>OMIM:615763</td>\n",
       "      <td>#615763 CORTICAL DYSPLASIA, COMPLEX, WITH OTHE...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HP:0002365</td>\n",
       "      <td>OMIM:615763</td>\n",
       "      <td>TAS</td>\n",
       "      <td>NaN</td>\n",
       "      <td>HP:0040283</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>P</td>\n",
       "      <td>HPO:skoehler[2014-08-24]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   #DatabaseID                                        DiseaseName Qualifier  \\\n",
       "0  OMIM:210100    BETA-AMINOISOBUTYRIC ACID, URINARY EXCRETION OF       NaN   \n",
       "1  OMIM:210100    BETA-AMINOISOBUTYRIC ACID, URINARY EXCRETION OF       NaN   \n",
       "2  OMIM:163600                                   NIPPLES INVERTED       NaN   \n",
       "3  OMIM:163600                                   NIPPLES INVERTED       NaN   \n",
       "4  OMIM:615763  #615763 CORTICAL DYSPLASIA, COMPLEX, WITH OTHE...       NaN   \n",
       "\n",
       "       HPO_ID    Reference Evidence Onset   Frequency  Sex Modifier Aspect  \\\n",
       "0  HP:0000007  OMIM:210100      IEA   NaN         NaN  NaN      NaN      I   \n",
       "1  HP:0003355  OMIM:210100      IEA   NaN         NaN  NaN      NaN      P   \n",
       "2  HP:0000006  OMIM:163600      IEA   NaN         NaN  NaN      NaN      I   \n",
       "3  HP:0003186  OMIM:163600      IEA   NaN         NaN  NaN      NaN      P   \n",
       "4  HP:0002365  OMIM:615763      TAS   NaN  HP:0040283  NaN      NaN      P   \n",
       "\n",
       "                Biocuration  \n",
       "0       HPO:iea[2009-02-17]  \n",
       "1  HPO:skoehler[2009-02-17]  \n",
       "2       HPO:iea[2009-02-17]  \n",
       "3       HPO:iea[2009-02-17]  \n",
       "4  HPO:skoehler[2014-08-24]  "
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_hpoa = pd.read_table(r'hpo/phenotype.hpoa', sep='\\t', header=4, encoding = \"ISO-8859-1\")\n",
    "print('Nr of HPO codes:', len(df_hpoa))\n",
    "df_hpoa.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Convert Entrez to Ensemble [ OLD FUNCTION ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#import sys\n",
    "#import mygene\n",
    "#def translateEnsembl(genes):\n",
    "#    l_val = []\n",
    "#    for gene in genes:\n",
    "#        result = mg.query(gene, scopes=\"symbol\", fields=[\"ensembl\"], species=\"human\", verbose=False)\n",
    "#        hgnc_name = gene\n",
    "#\n",
    "#        for hit in result[\"hits\"]:\n",
    "#            if \"ensembl\" in hit and \"gene\" in hit[\"ensembl\"]:\n",
    "#                print(\"%s\\t%s\\n\" % (hgnc_name, hit[\"ensembl\"][\"gene\"]))\n",
    "#                l_val.append(hit[\"ensembl\"][\"gene\"])\n",
    "#    return l_val\n",
    "#\n",
    "#\n",
    "#l_ensembl = translateEnsembl(genes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Employ PHRANK"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "from phrank.phrank import Phrank\n",
    "from phrank.phrank import utils as phrank_utils\n",
    "\n",
    "# hpo/pheno_2_gene.txt\n",
    "DAG=\"phrank/demo/data/hpodag.txt\"\n",
    "DISEASE_TO_PHENO=\"phrank/demo/data/disease_to_pheno.build127.txt\"\n",
    "DISEASE_TO_GENE=\"phrank/demo/data/gene_to_disease.build127.txt\"\n",
    "GENE_TO_PHENO=\"phrank/demo/data/gene_to_pheno.amelie.txt\"\n",
    "GENE_TO_PHENO='hpo/pheno_2_gene.txt'\n",
    "p_hpo = Phrank(DAG, geneannotationsfile=GENE_TO_PHENO) # diseaseannotationsfile=DISEASE_TO_PHENO, diseasegenefile=DISEASE_TO_GENE"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get assoc genes upon providing a phenotype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_assoc_genes(hpo_id, df_hpo, df_hpoa):\n",
    "    \"\"\"\n",
    "    Get genes that correspond to the provided phenotype (HPO)\n",
    "    \"\"\"\n",
    "    \n",
    "    output_genes = list(df_hpo[df_hpo['#Format: HPO-id']==hpo_id]['entrez-gene-symbol'].unique())\n",
    "    #l_diseases = list(df_hpo[df_hpo['#Format: HPO-id']==hpo_id]['disease-ID for link'].unique())\n",
    "    #output_diseases = list(df_hpoa[df_hpoa['Reference'].isin(l_diseases)]['DiseaseName'].unique())\n",
    "    #output_genes = list(df_hpo[df_hpo['Reference'].isin(l_diseases)]['entrez-gene-symbol'].unique())\n",
    "    # output_diseases, \n",
    "    #pd.DataFrame(output_genes,columns=['Gene'])\n",
    "    \n",
    "    return pd.DataFrame.from_dict({'Gene': output_genes, 'HPO' : [hpo_id for i in range(len(output_genes))]})\n",
    "\n",
    "#hpo_id = pheno_IARS[0] # hpo_list \n",
    "df_genes = pd.DataFrame()\n",
    "for hpo_id in pheno_IARS: # pheno_ACTG # pheno_IARS\n",
    "    df_genes2 = get_assoc_genes(hpo_id, df_hpo, df_hpoa)\n",
    "    df_genes = df_genes.append(df_genes2, ignore_index=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Defining patient genes and phenotypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "pheno_IARS = pheno_ACTG[-10:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'IARS1' in genes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [],
   "source": [
    "patient_phenotypes = pheno_IARS #  #  pheno_IARS\n",
    "genes = list(df_genes['Gene'].unique())\n",
    "patient_genes = set(genes) # genes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "101 10\n"
     ]
    }
   ],
   "source": [
    "print(len(genes), len(patient_phenotypes)) #  pheno_ACTG"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Rank genes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Gene ranking\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gene ID</th>\n",
       "      <th>Similarity Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>TGFB1</td>\n",
       "      <td>34.329135</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>COL7A1</td>\n",
       "      <td>28.194166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ATP6V0A2</td>\n",
       "      <td>28.127676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>FH</td>\n",
       "      <td>27.654747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>MMP1</td>\n",
       "      <td>27.354597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4197</th>\n",
       "      <td>ADAMTS18</td>\n",
       "      <td>0.001610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4198</th>\n",
       "      <td>ADAM9</td>\n",
       "      <td>0.001610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4199</th>\n",
       "      <td>ADAM22</td>\n",
       "      <td>0.001610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4200</th>\n",
       "      <td>ACE</td>\n",
       "      <td>0.001610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4201</th>\n",
       "      <td>ABCA7</td>\n",
       "      <td>0.001610</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4202 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       Gene ID  Similarity Score\n",
       "0        TGFB1         34.329135\n",
       "1       COL7A1         28.194166\n",
       "2     ATP6V0A2         28.127676\n",
       "3           FH         27.654747\n",
       "4         MMP1         27.354597\n",
       "...        ...               ...\n",
       "4197  ADAMTS18          0.001610\n",
       "4198     ADAM9          0.001610\n",
       "4199    ADAM22          0.001610\n",
       "4200       ACE          0.001610\n",
       "4201     ABCA7          0.001610\n",
       "\n",
       "[4202 rows x 2 columns]"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# sorting the genes by best match\n",
    "df_phrank = pd.DataFrame(columns=['Gene ID','Similarity Score'])\n",
    "\n",
    "gene_ranking = p_hpo.rank_genes(patient_genes, patient_phenotypes)\n",
    "\n",
    "print (\"\\nGene ranking\")\n",
    "for ix, gene_info in enumerate(gene_ranking):\n",
    "    #print (\"ensembl gene id: %s\\tsimilarity score: %.2f\"%(gene_info[1],gene_info[0]))\n",
    "    df_phrank.loc[ix] = [gene_info[1], gene_info[0]]\n",
    "df_phrank"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [],
   "source": [
    "#df_phrank.head(20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate How much better than Random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "51"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gene = 'COG7'\n",
    "\n",
    "col2hpo = columns_linked_to_hpo.column_to_hpo\n",
    "\n",
    "hpo_list = translateToHPO(pheno_ACTG, col2hpo)\n",
    "\n",
    "#[df_phrank['Gene ID']==gene]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random permutations\n",
    "- Provide 100 random phenotypic set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Gene ID</th>\n",
       "      <th>Similarity Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>VPS13A</td>\n",
       "      <td>23.370278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NSMCE2</td>\n",
       "      <td>19.361794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>TKFC</td>\n",
       "      <td>18.535729</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>TCF4</td>\n",
       "      <td>17.492435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SUCLG1</td>\n",
       "      <td>16.968603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>SLC2A1</td>\n",
       "      <td>16.652866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>GBA</td>\n",
       "      <td>16.563006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>TP53</td>\n",
       "      <td>15.469064</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>TERT</td>\n",
       "      <td>15.469064</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>SDHD</td>\n",
       "      <td>15.469064</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Gene ID  Similarity Score\n",
       "0  VPS13A         23.370278\n",
       "1  NSMCE2         19.361794\n",
       "2    TKFC         18.535729\n",
       "3    TCF4         17.492435\n",
       "4  SUCLG1         16.968603\n",
       "5  SLC2A1         16.652866\n",
       "6     GBA         16.563006\n",
       "7    TP53         15.469064\n",
       "8    TERT         15.469064\n",
       "9    SDHD         15.469064"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_phrank.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4202"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#hpo_list# = pheno_ACTG\n",
    "#patient_genes\n",
    "len(df_phrank)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Time elapsed on working...\n",
      "Iteration: 0\n",
      "Iteration: 5\n",
      "855\n",
      "Iteration: 10\n",
      "623\n",
      "Iteration: 15\n",
      "702\n",
      "Iteration: 20\n",
      "746\n",
      "Iteration: 25\n",
      "803\n",
      "Iteration: 30\n",
      "78\n",
      "Iteration: 35\n",
      "589\n",
      "Iteration: 40\n",
      "85\n",
      "Iteration: 45\n",
      "258\n",
      "Iteration: 50\n",
      "1070\n",
      "Iteration: 55\n",
      "65\n",
      "Iteration: 60\n",
      "320\n",
      "Iteration: 65\n",
      "104\n",
      "Iteration: 70\n",
      "866\n",
      "Iteration: 75\n",
      "528\n",
      "Iteration: 80\n",
      "184\n",
      "Iteration: 85\n",
      "620\n",
      "Iteration: 90\n",
      "354\n",
      "Iteration: 95\n",
      "99\n",
      "Time consumed in working:  4542.6522743701935\n"
     ]
    }
   ],
   "source": [
    "import random\n",
    "import time\n",
    "#imp.reload(func)\n",
    "\n",
    "start = time.time()\n",
    "print(\"Time elapsed on working...\")\n",
    "\n",
    "\n",
    "list_pos_genes = []\n",
    "\n",
    "gene = 'ACTB'\n",
    "\n",
    "SEED = 448\n",
    "random.seed(SEED)\n",
    "for i in range(100): \n",
    "    \n",
    "    \n",
    "    random_hpo_list = random.sample(pheno_ACTG, 10)\n",
    "    #print(i, random_hpo_list)\n",
    "    #continue \n",
    "    #hpo_id = pheno_IARS[0]\n",
    "    df_genes = pd.DataFrame()\n",
    "    #print(hpo_list)\n",
    "    for hpo_id in hpo_list: # pheno_ACTG # pheno_IARS\n",
    "        df_genes2 = get_assoc_genes(hpo_id, df_hpo, df_hpoa)\n",
    "        df_genes = df_genes.append(df_genes2, ignore_index=True)\n",
    "\n",
    "    ## Define all Input Parameters\n",
    "    #patient_phenotypes = hpo_list #  #  pheno_IARS\n",
    "    genes = list(df_genes['Gene'].unique())\n",
    "    patient_genes = set(genes) # genes\n",
    "\n",
    "    df_phrank = pd.DataFrame(columns=['Gene ID','Similarity Score'])\n",
    "\n",
    "    gene_ranking = p_hpo.rank_genes(patient_genes, random_hpo_list)\n",
    "    \n",
    "    if i % 5 == 0:\n",
    "        print('Iteration:', i)\n",
    "        if list_pos_genes != []:\n",
    "            print(list_pos_genes[-1])\n",
    "    #print (\"\\nGene ranking\")\n",
    "    for ix, gene_info in enumerate(gene_ranking):\n",
    "        #print (\"ensembl gene id: %s\\tsimilarity score: %.2f\"%(gene_info[1],gene_info[0]))\n",
    "        df_phrank.loc[ix] = [gene_info[1], gene_info[0]]\n",
    "        \n",
    "    try :\n",
    "        position_gene = df_phrank.index[df_phrank['Gene ID'] == gene].values[0]\n",
    "        list_pos_genes.append(position_gene)\n",
    "    except:  # not found in list -> then be gracious -> put it on the end of the list. (could be even lower)\n",
    "        list_pos_genes.append(len(df_phrank))\n",
    "    #df_phrank.head()\n",
    "end = time.time()\n",
    "print(\"Time consumed in working: \",end - start)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11\n"
     ]
    }
   ],
   "source": [
    "thresh=100\n",
    "l_within = [i <= thresh-1 for i in list_pos_genes]\n",
    "print(l_within.count(True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1628, 13, 550.31)"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "max(list_pos_genes), min(list_pos_genes), np.mean(list_pos_genes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'745,820,1270,142,855,294,657,197,1434,623,214,185,1099,611,702,701,537,740,95,746,1006,992,674,969,803,408,64,258,520,78,846,130,462,419,589,1628,184,156,51,85,556,115,439,116,258,158,97,1209,714,1070,416,557,418,321,65,264,274,256,1502,320,167,525,595,958,104,303,318,347,839,866,321,1133,42,883,528,538,795,718,468,184,478,1062,654,903,620,1332,241,269,85,354,327,1043,916,646,99,1234,903,13,368,738'"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "','.join([str(i) for i in list_pos_genes])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Position ACTB in Phrank list')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "\n",
    "ax = sns.boxplot(x=l_ACTB)\n",
    "ax.tick_params(labelsize=12)\n",
    "ax.set_xlabel(\"Position ACTB in Phrank list\",fontsize=15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_ACTB = [200,42,15,32,8,124,43,21,7,10,278,385,44,57,35,235,50,211,6,161,146,42,22,75,8,11,26,74,8,15,3,70,102,195,59,17,58,42,18,43,55,81,162,30,90,77,17,61,7,151,8,16,244,3,0,9,191,114,24,2,27,41,27,145,163,38,3,15,134,0,74,61,68,0,74,75,0,115,117,262,80,1,94,213,0,279,2,343,2,16,4,20,0,18,17,55,40,179,13,33]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "l_ACTG1 = [79,3,43,21,42,213,5,0,31,15,99,16,211,10,6,0,9,1,96,51,23,61,120,3,76,1,8,3,96,1,5,14,1,6,57,11,54,43,22,163,4,4,38,28,73,37,2,2,4,43,52,15,14,41,24,28,15,2,2,66,373,64,144,112,4,44,2,1,0,3,268,32,2,6,6,12,4,15,2,8,33,9,1,7,0,12,15,2,10,29,2,6,23,8,78,0,53,13,2,10]\n",
    "l_ACTB = [200,42,15,32,8,124,43,21,7,10,278,385,44,57,35,235,50,211,6,161,146,42,22,75,8,11,26,74,8,15,3,70,102,195,59,17,58,42,18,43,55,81,162,30,90,77,17,61,7,151,8,16,244,3,0,9,191,114,24,2,27,41,27,145,163,38,3,15,134,0,74,61,68,0,74,75,0,115,117,262,80,1,94,213,0,279,2,343,2,16,4,20,0,18,17,55,40,179,13,33]\n",
    "l_IARS = [745,820,1270,142,855,294,657,197,1434,623,214,185,1099,611,702,701,537,740,95,746,1006,992,674,969,803,408,64,258,520,78,846,130,462,419,589,1628,184,156,51,85,556,115,439,116,258,158,97,1209,714,1070,416,557,418,321,65,264,274,256,1502,320,167,525,595,958,104,303,318,347,839,866,321,1133,42,883,528,538,795,718,468,184,478,1062,654,903,620,1332,241,269,85,354,327,1043,916,646,99,1234,903,13,368,738]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot histogram ACTG1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([70., 16.,  6.,  3.,  1.,  2.,  0.,  1.,  0.,  1.]),\n",
       " array([  0. ,  37.3,  74.6, 111.9, 149.2, 186.5, 223.8, 261.1, 298.4,\n",
       "        335.7, 373. ]),\n",
       " <a list of 10 Patch objects>)"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "n_bins = 10\n",
    "\n",
    "\n",
    "plt.hist(list_pos_genes, bins=n_bins)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset #2 ACTG / ACTB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "non_null_columns = [col for col in df_actb.columns if (df_actb.loc[:, col]==1).any()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(92, 97)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(non_null_columns), len(df_actb.columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Category</th>\n",
       "      <th>Id</th>\n",
       "      <th>Protein</th>\n",
       "      <th>High frequency hearing loss</th>\n",
       "      <th>Short stature</th>\n",
       "      <th>microcephaly</th>\n",
       "      <th>hearing loss</th>\n",
       "      <th>ID</th>\n",
       "      <th>Developmental delay (Global)</th>\n",
       "      <th>Motor Delay</th>\n",
       "      <th>...</th>\n",
       "      <th>Recurrent infections</th>\n",
       "      <th>Photosensitivity</th>\n",
       "      <th>Polyarthralgia</th>\n",
       "      <th>Hepatomegaly</th>\n",
       "      <th>Hypothyroidism</th>\n",
       "      <th>Leukocytose</th>\n",
       "      <th>Trombocytopenie</th>\n",
       "      <th>Asthma</th>\n",
       "      <th>Strabismus</th>\n",
       "      <th>Growth hormone deficiency</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Gain Of Function</td>\n",
       "      <td>61456</td>\n",
       "      <td>p.Arg196Cys</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>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>Gain Of Function</td>\n",
       "      <td>58248</td>\n",
       "      <td>p.Arg196His</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>Gain Of Function</td>\n",
       "      <td>59169</td>\n",
       "      <td>p.Arg196His</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>3</th>\n",
       "      <td>Gain Of Function</td>\n",
       "      <td>LR04-173</td>\n",
       "      <td>p.Arg196His</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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>Gain Of Function</td>\n",
       "      <td>LR09-079</td>\n",
       "      <td>p.Arg196His</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</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 × 96 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           Category        Id      Protein  High frequency hearing loss  \\\n",
       "0  Gain Of Function     61456  p.Arg196Cys                            0   \n",
       "1  Gain Of Function     58248  p.Arg196His                            0   \n",
       "2  Gain Of Function     59169  p.Arg196His                            0   \n",
       "3  Gain Of Function  LR04-173  p.Arg196His                            0   \n",
       "4  Gain Of Function  LR09-079  p.Arg196His                            0   \n",
       "\n",
       "   Short stature   microcephaly   hearing loss    ID  \\\n",
       "0              0              0              0     0   \n",
       "1              0              0              0     1   \n",
       "2              1              1              0     1   \n",
       "3              1              1              1     1   \n",
       "4              1              1              0     1   \n",
       "\n",
       "   Developmental delay (Global)   Motor Delay  ...  Recurrent infections  \\\n",
       "0                             0             0  ...                     0   \n",
       "1                             1             0  ...                     0   \n",
       "2                             0             0  ...                     0   \n",
       "3                             0             0  ...                     0   \n",
       "4                             0             0  ...                     0   \n",
       "\n",
       "   Photosensitivity  Polyarthralgia  Hepatomegaly  Hypothyroidism  \\\n",
       "0               0.0             0.0           0.0               0   \n",
       "1               0.0             0.0           0.0               0   \n",
       "2               0.0             0.0           0.0               0   \n",
       "3               0.0             0.0           0.0               0   \n",
       "4               0.0             0.0           0.0               0   \n",
       "\n",
       "   Leukocytose  Trombocytopenie  Asthma  Strabismus  Growth hormone deficiency  \n",
       "0            0                0       0           0                          0  \n",
       "1            0                0       0           0                          0  \n",
       "2            0                0       0           0                          0  \n",
       "3            0                0       0           0                          0  \n",
       "4            0                0       0           0                          0  \n",
       "\n",
       "[5 rows x 96 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df_actb = pd.read_csv(r'data/ACTB/ACTB_data.csv', sep=',', header=0)\n",
    "df_actg1 = pd.read_csv(r'data/ACTG1/ACTG1_data.csv', sep=',', header=0)\n",
    "df_nat6 = pd.read_csv(r'data/NAT6/NAT6_data.csv', sep=',', header=0)\n",
    "df_actb.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_actg1['Category'] = df_actg1['Category'] + ' (ACTG1)'\n",
    "df_actb['Category'] = df_actb['Category'] + ' (ACTB)'\n",
    "df_actb['Origin'] = 'ACTB'\n",
    "df_actg1['Origin'] = 'ACTG1'\n",
    "#df_nat6['Origin'] = 'NAT6'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fix typo's and create new category names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "cols = list(df_actg1.loc[:, ~df_actg1.columns.isin(['Category', 'Id', 'Protein', 'Origin'])].columns)\n",
    "\n",
    "#df_all = df_actg1.copy() \n",
    "df_all = pd.concat([df_actg1, df_actb], ignore_index=True) # df_actg1.copy() \n",
    "df_all = pd.concat([df_all, df_nat6], ignore_index=True) # \n",
    "df_all = df_all.rename(columns={\"Retrognatia/micrognathia\" : \"Retrognathia/micrognathia\", \"Leukocytose\": \"Leukocytosis\", \"Trombocytopenie\": \"thrombocytopenia\", \"Other Psychiatric diagnosis (shizofrenia, depression, other, not autism, or ADHD/ADD)\" : \"Other Psychiatric diagnosis (schizophrenia, depression, other, not autism, or ADHD/ADD)\", \"Mid face hypoplasia\": \"Midface hypoplasia\"})\n",
    "df_all['Category'] = df_all['Category'].replace({'Other/Myosin binding (ACTG1)': 'Other (ACTG1)', 'Other/Fimbrin binding (ACTG1)': 'Other (ACTG1)', 'N-terminus (ACTG1)': 'Other (ACTG1)', 'N-terminus (ACTB)': 'Other (ACTB)'})\n",
    "\n",
    "df_pheno = df_all.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2.3 Translate headers\n",
    "In order to calculate the HRSS it is important to translate the headers to the corresponding HPO-codes!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "persistent embryonic sructures False\n",
      "persistent embryonic sructures False\n",
      "persistent embryonic sructures False\n",
      "polyarthralgia False\n",
      "persistent embryonic sructures False\n",
      "persistent embryonic sructures False\n"
     ]
    }
   ],
   "source": [
    "import columns_linked_to_hpo\n",
    "def infer_HPO(row, column_to_hpo):\n",
    "    hpo_list = []\n",
    "    cols = list(row.index)\n",
    "    for i in cols:\n",
    "        val = row[i]\n",
    "        #print(val)\n",
    "        if val == 1:\n",
    "            col = i.lower().strip()\n",
    "            if (col in column_to_hpo.keys()) == False :\n",
    "                print(col, col in column_to_hpo.keys())\n",
    "            if col in column_to_hpo.keys():\n",
    "                if type(column_to_hpo[col]) == list:\n",
    "                    hpo_list.extend(column_to_hpo[col])\n",
    "                else :\n",
    "                    hpo_list.append(column_to_hpo[col])\n",
    "    return hpo_list\n",
    "\n",
    "hp_patients = list(df_all.apply(lambda x: infer_HPO(x, columns_linked_to_hpo.column_to_hpo) , axis= 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save PhenoTable to file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('data/deepdata/actg_patpheno.csv', 'w') as f:\n",
    "    cnt = 0\n",
    "    for pat in df_all['Id']:\n",
    "        f.write(str(pat) + ',' + ','.join(hp_patients[cnt]) + '\\n')\n",
    "        cnt +=1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset #3 COG"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Index</th>\n",
       "      <th>Superclass</th>\n",
       "      <th>Clinical feature</th>\n",
       "      <th>HPO code</th>\n",
       "      <th>COG1</th>\n",
       "      <th>COG2</th>\n",
       "      <th>COG4</th>\n",
       "      <th>COG5</th>\n",
       "      <th>COG6</th>\n",
       "      <th>COG7</th>\n",
       "      <th>COG8</th>\n",
       "      <th>All</th>\n",
       "      <th>Ass. genes</th>\n",
       "      <th>Occ. ratio</th>\n",
       "      <th>Unnamed: 14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Abnormality of blood or blood-forming tissues</td>\n",
       "      <td>HP:0001871</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>17</td>\n",
       "      <td>944</td>\n",
       "      <td>0.018</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>Abnormal bleeding</td>\n",
       "      <td>HP:0001892</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>340</td>\n",
       "      <td>0.018</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>Gingival bleeding</td>\n",
       "      <td>HP:0000225</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>1</td>\n",
       "      <td>27</td>\n",
       "      <td>0.037</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>Epistaxis</td>\n",
       "      <td>HP:0000421</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>1</td>\n",
       "      <td>60</td>\n",
       "      <td>0.017</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>Spontaneous, recurrent epistaxis</td>\n",
       "      <td>HP:0004406</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>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0.250</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Index  Superclass                               Clinical feature  \\\n",
       "0      1           1  Abnormality of blood or blood-forming tissues   \n",
       "1      2           2                              Abnormal bleeding   \n",
       "2      3           3                              Gingival bleeding   \n",
       "3      4           3                                      Epistaxis   \n",
       "4      5           4               Spontaneous, recurrent epistaxis   \n",
       "\n",
       "     HPO code  COG1  COG2  COG4  COG5  COG6  COG7  COG8  All  Ass. genes  \\\n",
       "0  HP:0001871     1     1     2     1     9     2     1   17         944   \n",
       "1  HP:0001892     0     0     0     0     4     1     1    6         340   \n",
       "2  HP:0000225     0     0     0     0     1     0     0    1          27   \n",
       "3  HP:0000421     0     0     0     0     1     0     0    1          60   \n",
       "4  HP:0004406     0     0     0     0     1     0     0    1           4   \n",
       "\n",
       "   Occ. ratio  Unnamed: 14  \n",
       "0       0.018          NaN  \n",
       "1       0.018          NaN  \n",
       "2       0.037          NaN  \n",
       "3       0.017          NaN  \n",
       "4       0.250          NaN  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "df_cog_all = pd.read_csv(r'data/COG/COG.csv', sep=';', header=0)\n",
    "#df_nat6 = pd.read_csv(r'data/NAT6/NAT6_data.csv', sep=',', header=0)\n",
    "df_cog_all.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Option 1: all COGs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "hpo_list = df_cog_all.sort_values('Occ. ratio', ascending=False).head(10)['HPO code'].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Option 2: Per COG subgroup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "df_cog_split = pd.read_csv(r'data/COG/JMID_split_patients.csv', sep=';', header=0,index_col=0)\n",
    "df_cog_split = df_cog_split.apply(lambda x: x.str.replace(',','.'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_cog_split.sort_values('COG2', axis=1, ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "hpo_list = df_cog_split.sort_values('COG8', axis=1, ascending=False).columns[:10].values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "HPOcode\n",
       "COG1    0.0000011\n",
       "COG2      0.00106\n",
       "COG4      0.00212\n",
       "COG5       0.0011\n",
       "COG6       0.0095\n",
       "COG7       0.0021\n",
       "COG8       0.0011\n",
       "Name: HP:0001871, dtype: object"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_cog_split['HP:0001871']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Id</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.422868</td>\n",
       "      <td>0.155740</td>\n",
       "      <td>0.250735</td>\n",
       "      <td>0.103815</td>\n",
       "      <td>0.103815</td>\n",
       "      <td>0.096273</td>\n",
       "      <td>0.090208</td>\n",
       "      <td>0.131889</td>\n",
       "      <td>0.111612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.155740</td>\n",
       "      <td>0.386267</td>\n",
       "      <td>0.231097</td>\n",
       "      <td>0.108952</td>\n",
       "      <td>0.108952</td>\n",
       "      <td>0.102197</td>\n",
       "      <td>0.095203</td>\n",
       "      <td>0.136614</td>\n",
       "      <td>0.133293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0.250735</td>\n",
       "      <td>0.231097</td>\n",
       "      <td>0.431094</td>\n",
       "      <td>0.088773</td>\n",
       "      <td>0.088773</td>\n",
       "      <td>0.082023</td>\n",
       "      <td>0.075700</td>\n",
       "      <td>0.151009</td>\n",
       "      <td>0.185922</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.103815</td>\n",
       "      <td>0.108952</td>\n",
       "      <td>0.088773</td>\n",
       "      <td>0.254624</td>\n",
       "      <td>0.254624</td>\n",
       "      <td>0.214028</td>\n",
       "      <td>0.201439</td>\n",
       "      <td>0.075983</td>\n",
       "      <td>0.064385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>0.103815</td>\n",
       "      <td>0.108952</td>\n",
       "      <td>0.088773</td>\n",
       "      <td>0.254624</td>\n",
       "      <td>0.254624</td>\n",
       "      <td>0.214028</td>\n",
       "      <td>0.201439</td>\n",
       "      <td>0.075983</td>\n",
       "      <td>0.064385</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>0.096273</td>\n",
       "      <td>0.102197</td>\n",
       "      <td>0.082023</td>\n",
       "      <td>0.214028</td>\n",
       "      <td>0.214028</td>\n",
       "      <td>0.230756</td>\n",
       "      <td>0.222643</td>\n",
       "      <td>0.064439</td>\n",
       "      <td>0.053277</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>0.090208</td>\n",
       "      <td>0.095203</td>\n",
       "      <td>0.075700</td>\n",
       "      <td>0.201439</td>\n",
       "      <td>0.201439</td>\n",
       "      <td>0.222643</td>\n",
       "      <td>0.235013</td>\n",
       "      <td>0.060575</td>\n",
       "      <td>0.047861</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>0.131889</td>\n",
       "      <td>0.136614</td>\n",
       "      <td>0.151009</td>\n",
       "      <td>0.075983</td>\n",
       "      <td>0.075983</td>\n",
       "      <td>0.064439</td>\n",
       "      <td>0.060575</td>\n",
       "      <td>0.411942</td>\n",
       "      <td>0.116681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>0.111612</td>\n",
       "      <td>0.133293</td>\n",
       "      <td>0.185922</td>\n",
       "      <td>0.064385</td>\n",
       "      <td>0.064385</td>\n",
       "      <td>0.053277</td>\n",
       "      <td>0.047861</td>\n",
       "      <td>0.116681</td>\n",
       "      <td>0.461902</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Id         0         1         2         3         4         5         6  \\\n",
       "0   0  0.422868  0.155740  0.250735  0.103815  0.103815  0.096273  0.090208   \n",
       "1   1  0.155740  0.386267  0.231097  0.108952  0.108952  0.102197  0.095203   \n",
       "2   2  0.250735  0.231097  0.431094  0.088773  0.088773  0.082023  0.075700   \n",
       "3   3  0.103815  0.108952  0.088773  0.254624  0.254624  0.214028  0.201439   \n",
       "4   4  0.103815  0.108952  0.088773  0.254624  0.254624  0.214028  0.201439   \n",
       "5   5  0.096273  0.102197  0.082023  0.214028  0.214028  0.230756  0.222643   \n",
       "6   6  0.090208  0.095203  0.075700  0.201439  0.201439  0.222643  0.235013   \n",
       "7   7  0.131889  0.136614  0.151009  0.075983  0.075983  0.064439  0.060575   \n",
       "8   8  0.111612  0.133293  0.185922  0.064385  0.064385  0.053277  0.047861   \n",
       "\n",
       "          7         8  \n",
       "0  0.131889  0.111612  \n",
       "1  0.136614  0.133293  \n",
       "2  0.151009  0.185922  \n",
       "3  0.075983  0.064385  \n",
       "4  0.075983  0.064385  \n",
       "5  0.064439  0.053277  \n",
       "6  0.060575  0.047861  \n",
       "7  0.411942  0.116681  \n",
       "8  0.116681  0.461902  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_hrss = df_hrss.rename_axis('Id').reset_index()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "my_env",
   "language": "python",
   "name": "my_env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}