sumstats/manipulate_gwas_sumstats.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c0e1ab62",
   "metadata": {},
   "source": [
    "All links in this tutorial are copied from https://ctg.cncr.nl/software/summary_statistics\n",
    "This commands are tested on Ubuntu and MacOS system, but not on Windows.\n",
    "Note that you can manually download the data, instead of using 'os.system' calls in the following cell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f689f070",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "\n",
    "# download the data (execute this cell just once)\n",
    "if not os.path.exists(f'sumstats/SavageJansen_2018_intelligence_metaanalysis.txt'):\n",
    "    url = 'https://ctg.cncr.nl/documents/p1651/SavageJansen_IntMeta_sumstats.zip'\n",
    "    os.system(f'wget {url}')\n",
    "    os.system('unzip SavageJansen_IntMeta_sumstats.zip')\n",
    "\n",
    "if not os.path.exists(f'sumstats/Morningness_sumstats_Jansenetal.txt'):\n",
    "    url = 'https://ctg.cncr.nl/documents/p1651/Morningness_sumstats_Jansenetal.txt.gz'\n",
    "    os.system(f'wget {url}')\n",
    "    os.system('gunzip Morningness_sumstats_Jansenetal.txt.gz')\n",
    "    os.system('mv Morningness_sumstats_Jansenetal.txt sumstats/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8f2f863c",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "# read in the data\n",
    "cog = pd.read_csv('sumstats/SavageJansen_2018_intelligence_metaanalysis.txt', \n",
    "                  delim_whitespace=True,\n",
    "                  usecols=['SNP', 'CHR', 'POS', 'A1', 'A2', 'stdBeta', 'SE', 'P'])\n",
    "\n",
    "mor = pd.read_csv('sumstats/Morningness_sumstats_Jansenetal.txt',\n",
    "                  delim_whitespace=True,\n",
    "                  usecols=['SNP', 'CHR', 'BP', 'A1', 'A2', 'OR', 'SE', 'P'],\n",
    "                  low_memory=False)\n",
    "\n",
    "mor.drop_duplicates('SNP', keep='first', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "24540465",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9295118 rows\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>SNP</th>\n",
       "      <th>CHR</th>\n",
       "      <th>POS</th>\n",
       "      <th>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>stdBeta</th>\n",
       "      <th>SE</th>\n",
       "      <th>P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>rs12184267</td>\n",
       "      <td>1</td>\n",
       "      <td>715265</td>\n",
       "      <td>t</td>\n",
       "      <td>c</td>\n",
       "      <td>0.006887</td>\n",
       "      <td>0.007519</td>\n",
       "      <td>0.3598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rs12184277</td>\n",
       "      <td>1</td>\n",
       "      <td>715367</td>\n",
       "      <td>a</td>\n",
       "      <td>g</td>\n",
       "      <td>-0.004914</td>\n",
       "      <td>0.007492</td>\n",
       "      <td>0.5116</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>rs12184279</td>\n",
       "      <td>1</td>\n",
       "      <td>717485</td>\n",
       "      <td>a</td>\n",
       "      <td>c</td>\n",
       "      <td>0.007912</td>\n",
       "      <td>0.007535</td>\n",
       "      <td>0.2938</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>rs116801199</td>\n",
       "      <td>1</td>\n",
       "      <td>720381</td>\n",
       "      <td>t</td>\n",
       "      <td>g</td>\n",
       "      <td>0.002217</td>\n",
       "      <td>0.007391</td>\n",
       "      <td>0.7644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>rs12565286</td>\n",
       "      <td>1</td>\n",
       "      <td>721290</td>\n",
       "      <td>c</td>\n",
       "      <td>g</td>\n",
       "      <td>0.004174</td>\n",
       "      <td>0.007375</td>\n",
       "      <td>0.5711</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           SNP  CHR     POS A1 A2   stdBeta        SE       P\n",
       "0   rs12184267    1  715265  t  c  0.006887  0.007519  0.3598\n",
       "1   rs12184277    1  715367  a  g -0.004914  0.007492  0.5116\n",
       "2   rs12184279    1  717485  a  c  0.007912  0.007535  0.2938\n",
       "3  rs116801199    1  720381  t  g  0.002217  0.007391  0.7644\n",
       "4   rs12565286    1  721290  c  g  0.004174  0.007375  0.5711"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(len(cog), 'rows')\n",
    "cog.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9282d7e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10857998 rows\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>SNP</th>\n",
       "      <th>CHR</th>\n",
       "      <th>BP</th>\n",
       "      <th>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>OR</th>\n",
       "      <th>SE</th>\n",
       "      <th>P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>rs12184267</td>\n",
       "      <td>1</td>\n",
       "      <td>715265</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>-0.000773</td>\n",
       "      <td>0.006405</td>\n",
       "      <td>0.9039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rs12184277</td>\n",
       "      <td>1</td>\n",
       "      <td>715367</td>\n",
       "      <td>G</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.003762</td>\n",
       "      <td>0.006390</td>\n",
       "      <td>0.5560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>rs12184279</td>\n",
       "      <td>1</td>\n",
       "      <td>717485</td>\n",
       "      <td>A</td>\n",
       "      <td>C</td>\n",
       "      <td>-0.003244</td>\n",
       "      <td>0.006390</td>\n",
       "      <td>0.6117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>rs116801199</td>\n",
       "      <td>1</td>\n",
       "      <td>720381</td>\n",
       "      <td>T</td>\n",
       "      <td>G</td>\n",
       "      <td>-0.002460</td>\n",
       "      <td>0.006276</td>\n",
       "      <td>0.6952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>rs12565286</td>\n",
       "      <td>1</td>\n",
       "      <td>721290</td>\n",
       "      <td>C</td>\n",
       "      <td>G</td>\n",
       "      <td>-0.001113</td>\n",
       "      <td>0.006298</td>\n",
       "      <td>0.8598</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           SNP CHR      BP A1 A2        OR        SE       P\n",
       "0   rs12184267   1  715265  T  C -0.000773  0.006405  0.9039\n",
       "1   rs12184277   1  715367  G  A -0.003762  0.006390  0.5560\n",
       "2   rs12184279   1  717485  A  C -0.003244  0.006390  0.6117\n",
       "3  rs116801199   1  720381  T  G -0.002460  0.006276  0.6952\n",
       "4   rs12565286   1  721290  C  G -0.001113  0.006298  0.8598"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(len(mor), 'rows')\n",
    "mor.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "4cff67a7",
   "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>SNP</th>\n",
       "      <th>CHR</th>\n",
       "      <th>POS</th>\n",
       "      <th>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>stdBeta</th>\n",
       "      <th>SE</th>\n",
       "      <th>P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>71026</th>\n",
       "      <td>rs71638823</td>\n",
       "      <td>1</td>\n",
       "      <td>22374032</td>\n",
       "      <td>t</td>\n",
       "      <td>c</td>\n",
       "      <td>0.021719</td>\n",
       "      <td>0.003872</td>\n",
       "      <td>2.036000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>804577</th>\n",
       "      <td>rs7563013</td>\n",
       "      <td>2</td>\n",
       "      <td>23934087</td>\n",
       "      <td>a</td>\n",
       "      <td>g</td>\n",
       "      <td>-0.017043</td>\n",
       "      <td>0.003020</td>\n",
       "      <td>1.662000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1574353</th>\n",
       "      <td>rs73139272</td>\n",
       "      <td>3</td>\n",
       "      <td>16858468</td>\n",
       "      <td>t</td>\n",
       "      <td>g</td>\n",
       "      <td>-0.024838</td>\n",
       "      <td>0.004057</td>\n",
       "      <td>9.235000e-10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2180661</th>\n",
       "      <td>rs12646225</td>\n",
       "      <td>4</td>\n",
       "      <td>696848</td>\n",
       "      <td>t</td>\n",
       "      <td>c</td>\n",
       "      <td>0.025128</td>\n",
       "      <td>0.004215</td>\n",
       "      <td>2.505000e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2890416</th>\n",
       "      <td>rs17826816</td>\n",
       "      <td>5</td>\n",
       "      <td>7519298</td>\n",
       "      <td>a</td>\n",
       "      <td>g</td>\n",
       "      <td>0.018375</td>\n",
       "      <td>0.003257</td>\n",
       "      <td>1.677000e-08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                SNP  CHR       POS A1 A2   stdBeta        SE             P\n",
       "71026    rs71638823    1  22374032  t  c  0.021719  0.003872  2.036000e-08\n",
       "804577    rs7563013    2  23934087  a  g -0.017043  0.003020  1.662000e-08\n",
       "1574353  rs73139272    3  16858468  t  g -0.024838  0.004057  9.235000e-10\n",
       "2180661  rs12646225    4    696848  t  c  0.025128  0.004215  2.505000e-09\n",
       "2890416  rs17826816    5   7519298  a  g  0.018375  0.003257  1.677000e-08"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show first genome-wide significant variant on chromosomes 1..5\n",
    "cog[cog['P']<=5e-08].drop_duplicates('CHR', keep='first').head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "f8fb5c85",
   "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>SNP</th>\n",
       "      <th>CHR</th>\n",
       "      <th>BP</th>\n",
       "      <th>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>OR</th>\n",
       "      <th>SE</th>\n",
       "      <th>P</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>27767</th>\n",
       "      <td>rs76626729</td>\n",
       "      <td>1</td>\n",
       "      <td>7744559</td>\n",
       "      <td>A</td>\n",
       "      <td>G</td>\n",
       "      <td>0.07929</td>\n",
       "      <td>0.014260</td>\n",
       "      <td>2.664000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>864695</th>\n",
       "      <td>rs12711951</td>\n",
       "      <td>2</td>\n",
       "      <td>4638414</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>0.01438</td>\n",
       "      <td>0.002546</td>\n",
       "      <td>1.627000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1837186</th>\n",
       "      <td>rs13081924</td>\n",
       "      <td>3</td>\n",
       "      <td>14400077</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>-0.01897</td>\n",
       "      <td>0.003387</td>\n",
       "      <td>2.144000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2832233</th>\n",
       "      <td>rs28634184</td>\n",
       "      <td>4</td>\n",
       "      <td>62873419</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>-0.01421</td>\n",
       "      <td>0.002587</td>\n",
       "      <td>3.966000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3618634</th>\n",
       "      <td>rs9291813</td>\n",
       "      <td>5</td>\n",
       "      <td>63842133</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>-0.01462</td>\n",
       "      <td>0.002622</td>\n",
       "      <td>2.489000e-08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                SNP CHR        BP A1 A2       OR        SE             P\n",
       "27767    rs76626729   1   7744559  A  G  0.07929  0.014260  2.664000e-08\n",
       "864695   rs12711951   2   4638414  T  C  0.01438  0.002546  1.627000e-08\n",
       "1837186  rs13081924   3  14400077  T  C -0.01897  0.003387  2.144000e-08\n",
       "2832233  rs28634184   4  62873419  T  C -0.01421  0.002587  3.966000e-08\n",
       "3618634   rs9291813   5  63842133  T  C -0.01462  0.002622  2.489000e-08"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mor[mor['P']<=5e-08].drop_duplicates('CHR', keep='first').head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f05bf8b0",
   "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>SNP</th>\n",
       "      <th>CHR</th>\n",
       "      <th>POS</th>\n",
       "      <th>A1_COG</th>\n",
       "      <th>A2_COG</th>\n",
       "      <th>stdBeta</th>\n",
       "      <th>P_COG</th>\n",
       "      <th>A1_MOR</th>\n",
       "      <th>A2_MOR</th>\n",
       "      <th>OR</th>\n",
       "      <th>P_MOR</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>rs17824247</td>\n",
       "      <td>2</td>\n",
       "      <td>144152539</td>\n",
       "      <td>t</td>\n",
       "      <td>c</td>\n",
       "      <td>-0.018726</td>\n",
       "      <td>2.341000e-11</td>\n",
       "      <td>C</td>\n",
       "      <td>T</td>\n",
       "      <td>-0.01389</td>\n",
       "      <td>1.066000e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>rs62380364</td>\n",
       "      <td>5</td>\n",
       "      <td>88107337</td>\n",
       "      <td>a</td>\n",
       "      <td>c</td>\n",
       "      <td>-0.017593</td>\n",
       "      <td>2.459000e-10</td>\n",
       "      <td>A</td>\n",
       "      <td>C</td>\n",
       "      <td>0.01356</td>\n",
       "      <td>2.064000e-09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>rs10274077</td>\n",
       "      <td>7</td>\n",
       "      <td>133624271</td>\n",
       "      <td>t</td>\n",
       "      <td>c</td>\n",
       "      <td>-0.015904</td>\n",
       "      <td>4.751000e-08</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>0.01320</td>\n",
       "      <td>2.191000e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>rs146951100</td>\n",
       "      <td>22</td>\n",
       "      <td>42386487</td>\n",
       "      <td>t</td>\n",
       "      <td>c</td>\n",
       "      <td>-0.028071</td>\n",
       "      <td>1.141000e-08</td>\n",
       "      <td>C</td>\n",
       "      <td>T</td>\n",
       "      <td>-0.02111</td>\n",
       "      <td>3.670000e-08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            SNP  CHR        POS A1_COG A2_COG   stdBeta         P_COG A1_MOR  \\\n",
       "0    rs17824247    2  144152539      t      c -0.018726  2.341000e-11      C   \n",
       "24   rs62380364    5   88107337      a      c -0.017593  2.459000e-10      A   \n",
       "45   rs10274077    7  133624271      t      c -0.015904  4.751000e-08      T   \n",
       "90  rs146951100   22   42386487      t      c -0.028071  1.141000e-08      C   \n",
       "\n",
       "   A2_MOR       OR         P_MOR  \n",
       "0       T -0.01389  1.066000e-09  \n",
       "24      C  0.01356  2.064000e-09  \n",
       "45      C  0.01320  2.191000e-08  \n",
       "90      T -0.02111  3.670000e-08  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show frst variant on each chromosome that is genome-wide significant for both traits\n",
    "cog_mor = pd.merge(cog[cog['P']<=5e-08][['SNP', 'CHR', 'POS', 'A1', 'A2', 'stdBeta', 'P']],\n",
    "                   mor[mor['P']<=5e-08][['SNP',               'A1', 'A2', 'OR', 'P']], \n",
    "                   on='SNP', how='inner', suffixes=('_COG', '_MOR'))\n",
    "cog_mor.drop_duplicates('CHR', keep='first')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "34f0b3aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "plt.figure(figsize=(15,5)); plt.subplot(1,2,1)\n",
    "plt.plot(-np.log10(np.array(range(1,1+len(cog)))/len(cog)), \n",
    "         -np.log10(sorted(cog['P'].values)), label='COG')\n",
    "plt.plot([0, 7], [0, 7], 'k--')\n",
    "plt.title('Q-Q plot for COG', fontsize=16);\n",
    "plt.xlabel('log10(p-value), expected under null', fontsize=16);\n",
    "plt.ylabel('log10(p-value), actual', fontsize=16);\n",
    "plt.legend()\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "plt.hist(cog['P'].values, density=True);\n",
    "plt.xlabel('PVALUE (deciles)')\n",
    "plt.ylabel('fraction in decile')\n",
    "plt.title('PVALUE histogram for COG', fontsize=16);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "58e44e6c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}