stitch_to_umls.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pandas import DataFrame, merge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "SIDER = DataFrame.from_csv('./temp/sider/meddra_all_se.tsv', sep='\\t', index_col=None, header=None)\n",
    "SIDER = SIDER.rename(columns={0: 'stitch_flat', 1: 'stitch_stereo',\n",
    "                              2: 'umls_label', 3: 'meddra',\n",
    "                              4: 'umls_meddra', 5: 'umls_def'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "SIDER_freq = DataFrame.from_csv('./temp/sider/meddra_freq.tsv', sep='\\t', index_col=None, header=None)\n",
    "SIDER_freq = SIDER_freq.rename(columns={0: 'stitch_flat', 1: 'stitch_stereo',\n",
    "                                        2: 'umls_label', 3: 'placebo',\n",
    "                                        4: 'average', 5: 'lower',\n",
    "                                        6: 'higher', 7: 'meddra',\n",
    "                                        8: 'umls_meddra', 9: 'umls_def'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "SIDER_complete = merge(SIDER, SIDER_freq, on=['stitch_flat', 'stitch_stereo',\n",
    "                                              'umls_label', 'meddra',\n",
    "                                              'umls_meddra', 'umls_def'], how='outer')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "309849 291632 479075\n"
     ]
    }
   ],
   "source": [
    "print(len(SIDER), len(SIDER_freq), len(SIDER_complete))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>stitch_flat</th>\n",
       "      <th>stitch_stereo</th>\n",
       "      <th>umls_label</th>\n",
       "      <th>placebo</th>\n",
       "      <th>average</th>\n",
       "      <th>lower</th>\n",
       "      <th>higher</th>\n",
       "      <th>meddra</th>\n",
       "      <th>umls_meddra</th>\n",
       "      <th>umls_def</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21%</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.21</td>\n",
       "      <td>LLT</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>Abdominal pain</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21%</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.21</td>\n",
       "      <td>PT</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>Abdominal pain</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>stitch_flat</th>\n",
       "      <th>stitch_stereo</th>\n",
       "      <th>umls_label</th>\n",
       "      <th>placebo</th>\n",
       "      <th>average</th>\n",
       "      <th>lower</th>\n",
       "      <th>higher</th>\n",
       "      <th>meddra</th>\n",
       "      <th>umls_meddra</th>\n",
       "      <th>umls_def</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21%</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.21</td>\n",
       "      <td>LLT</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>Abdominal pain</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21%</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.21</td>\n",
       "      <td>PT</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>Abdominal pain</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SIDER_freq[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>stitch_flat</th>\n",
       "      <th>stitch_stereo</th>\n",
       "      <th>umls_label</th>\n",
       "      <th>meddra</th>\n",
       "      <th>umls_meddra</th>\n",
       "      <th>umls_def</th>\n",
       "      <th>placebo</th>\n",
       "      <th>average</th>\n",
       "      <th>lower</th>\n",
       "      <th>higher</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000729</td>\n",
       "      <td>LLT</td>\n",
       "      <td>C0000729</td>\n",
       "      <td>Abdominal cramps</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000729</td>\n",
       "      <td>PT</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>Abdominal pain</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>stitch_flat</th>\n",
       "      <th>stitch_stereo</th>\n",
       "      <th>umls_label</th>\n",
       "      <th>meddra</th>\n",
       "      <th>umls_meddra</th>\n",
       "      <th>umls_def</th>\n",
       "      <th>placebo</th>\n",
       "      <th>average</th>\n",
       "      <th>lower</th>\n",
       "      <th>higher</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000729</td>\n",
       "      <td>LLT</td>\n",
       "      <td>C0000729</td>\n",
       "      <td>Abdominal cramps</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CID100000085</td>\n",
       "      <td>CID000010917</td>\n",
       "      <td>C0000729</td>\n",
       "      <td>PT</td>\n",
       "      <td>C0000737</td>\n",
       "      <td>Abdominal pain</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SIDER_complete[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "SIDER_complete['stitch_flat'] = SIDER_complete['stitch_flat'].map(lambda x:  x[:3] + 'm' + x[4:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "257048\n163711\n163224\n"
     ]
    }
   ],
   "source": [
    "SIDER_label = SIDER_complete[['stitch_flat', 'umls_label', 'lower', 'higher']].drop_duplicates()\n",
    "SIDER_label = SIDER_label.rename(columns={'stitch_flat': 'stitch_id', 'umls_label': 'umls_id'})\n",
    "\n",
    "SIDER_meddra = SIDER_complete[['stitch_flat', 'umls_meddra', 'lower', 'higher']].drop_duplicates()\n",
    "SIDER_meddra = SIDER_meddra.rename(columns={'stitch_flat': 'stitch_id', 'umls_meddra': 'umls_id'})\n",
    "    \n",
    "\n",
    "SIDER_union = merge(SIDER_meddra, SIDER_label, on=['stitch_id', 'umls_id', 'lower', 'higher'], how='outer')\n",
    "print(len(SIDER_union))\n",
    "SIDER_union = SIDER_union.drop_duplicates(subset=['stitch_id', 'umls_id'])\n",
    "print(len(SIDER_union))\n",
    "SIDER_union = SIDER_union[SIDER_union.stitch_id.notnull()]\n",
    "SIDER_union = SIDER_union[SIDER_union.umls_id.notnull()]\n",
    "print(len(SIDER_union))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sqlalchemy import create_engine\n",
    "\n",
    "engine = create_engine('postgresql://postgres:postgres@localhost:32768/efktr')\n",
    "SIDER_union.to_sql('stitch_to_umls', engine, index=False, if_exists='append')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    ""
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2.0
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}