NB2_analyze_DNV_overrepresentation_210419.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# DNV overrepresentation in network proximity to established disease genes\n",
    "\n",
    "- First run network propagation from each high confidence set, on each interactome (use netprop_zscore.py  - we ran it on AWS).  Requires graphs in gpickle format, and high confidence gene sets.\n",
    "\n",
    "- This notebook calculates dDNV over-representation in networks proximal to established disease genes\n",
    "\n",
    "-------------------\n",
    "\n",
    "Author: Brin Rosenthal (sbrosenthal@ucsd.edu)\n",
    "\n",
    "-------------------\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import networkx as nx\n",
    "import pandas as pd\n",
    "import random\n",
    "\n",
    "\n",
    "# latex rendering of text in graphs\n",
    "import matplotlib as mpl\n",
    "mpl.rc('text', usetex = False)\n",
    "mpl.rc('font', family = 'serif')\n",
    "\n",
    "from matplotlib import rcParams\n",
    "rcParams['font.family'] = 'sans-serif'\n",
    "rcParams['font.sans-serif'] = ['Arial']\n",
    "\n",
    "sns.set_style('white')\n",
    "sns.set_style(\"ticks\", {\"xtick.major.size\": 15, \"ytick.major.size\": 15})\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "import sys\n",
    "\n",
    "% matplotlib inline\n",
    "\n",
    "# for better labeling\n",
    "import adjustText"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load ASD, CHD genesets\n",
    "\n",
    "ASD HC genes come from table 4 in http://www.cell.com/neuron/fulltext/S0896-6273(15)00773-4 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "65"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ------- ASD high confidence genes (established disease genes) -------\n",
    "\n",
    "HC_genes_temp = pd.read_csv('data/HC_genes/ASD_HC.tsv',sep='\\t',index_col='Unnamed: 0')\n",
    "ASD_HC = [str(g[1:-1]).strip(\"'\") for g in HC_genes_temp['seed_genes'].tolist()[0][1:-1].split(', ')]\n",
    "\n",
    "len(ASD_HC)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "66"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# ------- High confidence CHD genes (established disease genes) -------\n",
    "\n",
    "# Load congenital heart defect recurring DNVs (from https://www.nature.com/articles/ng.3970)\n",
    "\n",
    "HC_genes_temp = pd.read_csv('data/HC_genes/CHD_HC.tsv',sep='\\t',index_col='Unnamed: 0')\n",
    "CHD_HC = [str(g[1:-1]).strip(\"'\") for g in HC_genes_temp['seed_genes'].tolist()[0][1:-1].split(', ')]\n",
    "\n",
    "len(CHD_HC)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number patients with a damaging ASD DNV (not in seed gene set):\n",
      "630\n",
      "number damaging ASD DNV (in patients which also have no seed gene variants):\n",
      "[u'A2ML1', u'AASS', u'ABCA1', u'ABCA13', u'ABCA7', u'ABCB9', u'ABCG1', u'ABI2', u'ABL1', u'ABL2']\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>Study</th>\n",
       "      <th>CHR</th>\n",
       "      <th>POS</th>\n",
       "      <th>REF</th>\n",
       "      <th>ALT</th>\n",
       "      <th>Proband ID</th>\n",
       "      <th>Class</th>\n",
       "      <th>Class (2)</th>\n",
       "      <th>Gene</th>\n",
       "      <th>Ensembl Gene ID</th>\n",
       "      <th>AA1</th>\n",
       "      <th>AA2</th>\n",
       "      <th>dbNSFP Meta-SVM Rankscore</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ASC</td>\n",
       "      <td>1</td>\n",
       "      <td>897739</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>09C87513</td>\n",
       "      <td>missense</td>\n",
       "      <td>D-mis</td>\n",
       "      <td>KLHL17</td>\n",
       "      <td>ENSG00000187961</td>\n",
       "      <td>L</td>\n",
       "      <td>P</td>\n",
       "      <td>0.87685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SSC</td>\n",
       "      <td>1</td>\n",
       "      <td>1222518</td>\n",
       "      <td>C</td>\n",
       "      <td>A</td>\n",
       "      <td>11563.p1</td>\n",
       "      <td>nonsense</td>\n",
       "      <td>lof</td>\n",
       "      <td>SCNN1D</td>\n",
       "      <td>ENSG00000162572</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ASC</td>\n",
       "      <td>1</td>\n",
       "      <td>2234766</td>\n",
       "      <td>C</td>\n",
       "      <td>T</td>\n",
       "      <td>08C78919</td>\n",
       "      <td>nonsense</td>\n",
       "      <td>lof</td>\n",
       "      <td>SKI</td>\n",
       "      <td>ENSG00000157933</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SSC</td>\n",
       "      <td>1</td>\n",
       "      <td>3519049</td>\n",
       "      <td>AC</td>\n",
       "      <td>A</td>\n",
       "      <td>14075.p1</td>\n",
       "      <td>frameshift</td>\n",
       "      <td>lof</td>\n",
       "      <td>MEGF6</td>\n",
       "      <td>ENSG00000162591</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SSC</td>\n",
       "      <td>1</td>\n",
       "      <td>11205057</td>\n",
       "      <td>T</td>\n",
       "      <td>C</td>\n",
       "      <td>11104.p1</td>\n",
       "      <td>missense</td>\n",
       "      <td>D-mis</td>\n",
       "      <td>MTOR</td>\n",
       "      <td>ENSG00000198793</td>\n",
       "      <td>M</td>\n",
       "      <td>V</td>\n",
       "      <td>0.83852</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Study CHR       POS REF ALT Proband ID       Class Class (2)    Gene  \\\n",
       "0   ASC   1    897739   T   C   09C87513    missense     D-mis  KLHL17   \n",
       "1   SSC   1   1222518   C   A   11563.p1    nonsense       lof  SCNN1D   \n",
       "2   ASC   1   2234766   C   T   08C78919    nonsense       lof     SKI   \n",
       "3   SSC   1   3519049  AC   A   14075.p1  frameshift       lof   MEGF6   \n",
       "4   SSC   1  11205057   T   C   11104.p1    missense     D-mis    MTOR   \n",
       "\n",
       "   Ensembl Gene ID  AA1  AA2  dbNSFP Meta-SVM Rankscore  \n",
       "0  ENSG00000187961    L    P                    0.87685  \n",
       "1  ENSG00000162572  NaN  NaN                        NaN  \n",
       "2  ENSG00000157933  NaN  NaN                        NaN  \n",
       "3  ENSG00000162591  NaN  NaN                        NaN  \n",
       "4  ENSG00000198793    M    V                    0.83852  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load all NDD DNVs (from 2015 CHD paper), extract SSC DNVs (autism)\n",
    "NDD_df = pd.read_excel('data/SSC/homsy_database_S08.xlsx',skiprows=1)\n",
    "\n",
    "# ------------ Update 4/11 ------------\n",
    "# Exclude patients with variants in the HC genes (Trey's comment)\n",
    "NDD_df = NDD_df[~NDD_df['Gene'].isin(ASD_HC+CHD_HC)]\n",
    "# ------------------------------------\n",
    "\n",
    "ASD_DNV = NDD_df[NDD_df['Study']=='SSC']\n",
    "print('number patients with a damaging ASD DNV (not in seed gene set):')\n",
    "print(len(ASD_DNV))\n",
    "ASD_DNV_VC = ASD_DNV['Gene'].value_counts() # count recurrences\n",
    "ASD_DNV = list(np.unique(ASD_DNV['Gene'].tolist()))\n",
    "\n",
    "\n",
    "print('number damaging ASD DNV (in patients which also have no seed gene variants):')\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "print(ASD_DNV[0:10])\n",
    "NDD_df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32\n",
      "[]\n"
     ]
    }
   ],
   "source": [
    "print(sum(ASD_DNV_VC>1)) # number recurrent ASD -- no seeds\n",
    "\n",
    "print(np.intersect1d(ASD_DNV_VC[ASD_DNV_VC>1].index.tolist(),ASD_HC)) # make sure there are no seeds in the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/pandas/util/_decorators.py:188: FutureWarning: The `sheetname` keyword is deprecated, use `sheet_name` instead\n",
      "  return func(*args, **kwargs)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2990\n",
      "2760\n",
      "number damaging CHD DNVs:\n",
      "629\n",
      "\n",
      "number recurrent damaging CHD DNVs:\n",
      "0\n",
      "\n",
      "number damaging CHD DNVs with NDD:\n",
      "155\n",
      "\n",
      "number damaging unique CHD DNVs with NDD:\n",
      "155\n",
      "\n",
      "number damaging CHD DNVs without NDD:\n",
      "270\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/pandas/core/computation/expressions.py:183: UserWarning: evaluating in Python space because the '*' operator is not supported by numexpr for the bool dtype, use '&' instead\n",
      "  .format(op=op_str, alt_op=unsupported[op_str]))\n"
     ]
    }
   ],
   "source": [
    "# Load all CHD DNVs (from 2017 paper https://www.nature.com/articles/ng.3970)\n",
    "\n",
    "CHD_DNV_df = pd.read_excel('data/CHD_databases_2017_paper/ng.3970-S3.xlsx',sheetname='S9',skiprows=1)\n",
    "CHD_DNV_df.index=CHD_DNV_df['Blinded ID']\n",
    "\n",
    "# exclude patients which have variants in HC genes -- from Trey's comment April 2019\n",
    "print(len(CHD_DNV_df))\n",
    "CHD_DNV_df = CHD_DNV_df[~CHD_DNV_df['Gene'].isin(ASD_HC+CHD_HC)]\n",
    "print(len(CHD_DNV_df))\n",
    "\n",
    "# apply the same filtering as for old data\n",
    "CHD_DNV = CHD_DNV_df[(CHD_DNV_df['Variant_Class']!='syn')*(CHD_DNV_df['Variant_Class']!='mis')]\n",
    "print('number damaging CHD DNVs:')\n",
    "print(len(CHD_DNV))\n",
    "\n",
    "CHD_DNV_VC = CHD_DNV['Gene'].value_counts() # count recurrences-- should be 0 because we removed them\n",
    "print('\\nnumber recurrent damaging CHD DNVs:')\n",
    "print(sum(CHD_DNV_VC>1))\n",
    "\n",
    "DNV_withNDD = CHD_DNV[CHD_DNV['NDD']=='Yes']\n",
    "print('\\nnumber damaging CHD DNVs with NDD:')\n",
    "print(len(DNV_withNDD)) # number of patients\n",
    "DNV_withNDD = DNV_withNDD['Gene'].unique().tolist()\n",
    "print('\\nnumber damaging unique CHD DNVs with NDD:')\n",
    "print(len(DNV_withNDD)) # number of genes\n",
    "\n",
    "DNV_noNDD = CHD_DNV[CHD_DNV['NDD']=='No']\n",
    "DNV_noNDD = DNV_noNDD['Gene'].unique().tolist()\n",
    "print('\\nnumber damaging CHD DNVs without NDD:')\n",
    "print(len(DNV_noNDD))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "629\n"
     ]
    }
   ],
   "source": [
    "CHD_DNV = list(np.unique(CHD_DNV['Gene']))\n",
    "print(len(CHD_DNV))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1700\n",
      "574\n"
     ]
    }
   ],
   "source": [
    "print(len(CHD_DNV_df['Blinded ID'].unique())) # num in CHD cohort\n",
    "print(sum(CHD_DNV_df['NDD']=='Yes'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "54\n"
     ]
    }
   ],
   "source": [
    "print(len(np.intersect1d(CHD_DNV,ASD_DNV)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load the control DNVs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number control DNVs:\n",
      "193\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>Blinded ID</th>\n",
       "      <th>CHROM</th>\n",
       "      <th>POS</th>\n",
       "      <th>REF</th>\n",
       "      <th>ALT</th>\n",
       "      <th>Gene</th>\n",
       "      <th>Ensembl Gene ID</th>\n",
       "      <th>Variant Class</th>\n",
       "      <th>AA Change</th>\n",
       "      <th>dbNSFP Meta-SVM Rankscore</th>\n",
       "      <th>Comments</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Blinded ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>11003.s1</th>\n",
       "      <td>11003.s1</td>\n",
       "      <td>1</td>\n",
       "      <td>202715044</td>\n",
       "      <td>G</td>\n",
       "      <td>C</td>\n",
       "      <td>KDM5B</td>\n",
       "      <td>ENSG00000117139</td>\n",
       "      <td>Nonsense</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11003.s1</th>\n",
       "      <td>11003.s1</td>\n",
       "      <td>5</td>\n",
       "      <td>178139609</td>\n",
       "      <td>G</td>\n",
       "      <td>A</td>\n",
       "      <td>ZNF354A</td>\n",
       "      <td>ENSG00000169131</td>\n",
       "      <td>Nonsense</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11003.s1</th>\n",
       "      <td>11003.s1</td>\n",
       "      <td>16</td>\n",
       "      <td>70819649</td>\n",
       "      <td>C</td>\n",
       "      <td>A</td>\n",
       "      <td>VAC14</td>\n",
       "      <td>ENSG00000103043</td>\n",
       "      <td>Missense</td>\n",
       "      <td>A127S</td>\n",
       "      <td>0.08066</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11003.s1</th>\n",
       "      <td>11003.s1</td>\n",
       "      <td>X</td>\n",
       "      <td>48374469</td>\n",
       "      <td>C</td>\n",
       "      <td>T</td>\n",
       "      <td>PORCN</td>\n",
       "      <td>ENSG00000102312</td>\n",
       "      <td>Missense</td>\n",
       "      <td>R288W</td>\n",
       "      <td>0.65878</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11008.s1</th>\n",
       "      <td>11008.s1</td>\n",
       "      <td>12</td>\n",
       "      <td>94769688</td>\n",
       "      <td>G</td>\n",
       "      <td>A</td>\n",
       "      <td>CCDC41</td>\n",
       "      <td>ENSG00000173588</td>\n",
       "      <td>Nonsense</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           Blinded ID CHROM        POS REF ALT     Gene  Ensembl Gene ID  \\\n",
       "Blinded ID                                                                 \n",
       "11003.s1     11003.s1     1  202715044   G   C    KDM5B  ENSG00000117139   \n",
       "11003.s1     11003.s1     5  178139609   G   A  ZNF354A  ENSG00000169131   \n",
       "11003.s1     11003.s1    16   70819649   C   A    VAC14  ENSG00000103043   \n",
       "11003.s1     11003.s1     X   48374469   C   T    PORCN  ENSG00000102312   \n",
       "11008.s1     11008.s1    12   94769688   G   A   CCDC41  ENSG00000173588   \n",
       "\n",
       "           Variant Class AA Change dbNSFP Meta-SVM Rankscore Comments  \n",
       "Blinded ID                                                             \n",
       "11003.s1        Nonsense       NaN                       NaN      NaN  \n",
       "11003.s1        Nonsense       NaN                       NaN      NaN  \n",
       "11003.s1        Missense     A127S                   0.08066      NaN  \n",
       "11003.s1        Missense     R288W                   0.65878      NaN  \n",
       "11008.s1        Nonsense       NaN                       NaN      NaN  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Same in new version \n",
    "\n",
    "# Load all Control DNVs (from Database S3 in http://science.sciencemag.org/content/sci/350/6265/1262.full.pdf)\n",
    "ctrl_DNV_df = pd.read_excel('data/SSC/homsy_database_S03.xlsx',skiprows=1)\n",
    "# exclude synonymous and non-damaging missense\n",
    "ctrl_DNV = ctrl_DNV_df[(ctrl_DNV_df['Variant Class']!='Synonymous')*(ctrl_DNV_df['Variant Class']!='Missense')] \n",
    "ctrl_DNV_VC = ctrl_DNV['Gene'].value_counts() # count recurrences\n",
    "ctrl_DNV = list(np.unique(ctrl_DNV['Gene'].tolist()))\n",
    "ctrl_DNV_df.index=ctrl_DNV_df['Blinded ID']\n",
    "print('number control DNVs:')\n",
    "print(len(ctrl_DNV))\n",
    "ctrl_DNV_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Define gene sets for rest of analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number ASD high confidence genes:\n",
      "65\n",
      "\n",
      "number CHD high confidence genes:\n",
      "66\n",
      "\n",
      "number shared high confidence genes\n",
      "5\n"
     ]
    }
   ],
   "source": [
    "# pick out which gene sets to investigate\n",
    "genes_ASD = ASD_HC\n",
    "print('number ASD high confidence genes:')\n",
    "print(len(genes_ASD))\n",
    "genes_CHD = CHD_HC\n",
    "print('\\nnumber CHD high confidence genes:')\n",
    "print(len(genes_CHD))\n",
    "\n",
    "print('\\nnumber shared high confidence genes')\n",
    "print(len(np.intersect1d(ASD_HC,CHD_HC)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load previously computed network prop z-scores, and define and load interactome\n",
    "\n",
    "computed with the following commands:\n",
    "\n",
    "python netprop_zscore.py 5000 /path/to/ASD_HC.tsv /path/to/G_PCnet.gpickle ASD degree_binning single  \n",
    "\n",
    "python netprop_zscore.py 5000 /path/to/ASD_HC.tsv /path/to/G_PCnet.gpickle ASD degree_binning single\n",
    "\n",
    "--> runtime ~ 12 hours each, output files included here\n",
    "\n",
    "PCnet downloaded from ndex and parsed to networkx format \n",
    "https://ndexbio.org/#/network/f93f402c-86d4-11e7-a10d-0ac135e8bacf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# load previously computed z-scores and focal interactome, \n",
    "import scipy\n",
    "\n",
    "num_reps = 5000\n",
    "focal_interactome='PCnet'\n",
    "rand_method = 'degree_binning' # can be 'degree_binning' or 'degree_ks_test' (deprecated)\n",
    "\n",
    "z_ASD = pd.read_csv('data/z_score_results/z_ASD_'+str(num_reps)+'_reps'+focal_interactome+'_'+rand_method+'.tsv',\n",
    "                sep='\\t',names=['gene','zscore'],index_col='gene')\n",
    "z_CHD = pd.read_csv('data/z_score_results/z_CHD_'+str(num_reps)+'_reps'+focal_interactome+'_'+rand_method+'.tsv',\n",
    "                       sep='\\t',names=['gene','zscore'],index_col='gene')\n",
    "\n",
    "z_ASD = z_ASD['zscore'].copy()\n",
    "z_CHD = z_CHD['zscore'].copy()\n",
    "\n",
    "\n",
    "                    \n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# reload the focal interactome (used to have more options... now we just use PCnet)\n",
    "if focal_interactome=='PCnet':\n",
    "    Gint = nx.read_gpickle('/Users/brinrosenthal/Documents/CCBB_tickets_data/PCnet/G_PCnet.gpickle')\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number edges in interactome:\n",
      "2724724\n",
      "\n",
      "number nodes in interactome\n",
      "19781\n"
     ]
    }
   ],
   "source": [
    "print('number edges in interactome:')\n",
    "print(len(Gint.edges()))\n",
    "print('\\nnumber nodes in interactome')\n",
    "print(len(Gint.nodes()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot location of DNVs in single z-score distributions, calculate KS statistic between DNV z-scores and full z-scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gene\n",
      "MED13L    4.215628\n",
      "PBRM1     4.207929\n",
      "VPS13D    4.089686\n",
      "PNPLA7    3.975127\n",
      "SETD2     3.852171\n",
      "Name: zscore, dtype: float64\n",
      "Ks_2sampResult(statistic=0.17767295597484278, pvalue=4.088998632167451e-16)\n",
      "gene\n",
      "VSIG1    7.312675\n",
      "CNTN2    4.981746\n",
      "RYR2     3.770345\n",
      "ITPR3    3.681369\n",
      "TRPC5    3.616511\n",
      "Name: zscore, dtype: float64\n",
      "Ks_2sampResult(statistic=0.16704716895244415, pvalue=3.1039089691566306e-15)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:8: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n",
      "  \n",
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:13: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n",
      "  del sys.path[0]\n",
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:25: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n",
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:39: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n",
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:42: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n",
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:56: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0, 0.43)"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 165.6x110.88 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 165.6x110.88 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from scipy.stats import ks_2samp\n",
    "plt.figure(figsize=(2.3,1.54))\n",
    "# plt.figure()\n",
    "sns.distplot(z_ASD.loc[np.setdiff1d(z_ASD.index.tolist(),ASD_HC)].dropna(),label='null model',color='#7E7E7E',kde=True,\n",
    "            hist_kws={'histtype':'stepfilled','lw':0},\n",
    "            kde_kws={'lw':.5})\n",
    "ASD_DNV_no_seeds = list(np.setdiff1d(ASD_DNV,ASD_HC))\n",
    "sns.distplot(z_ASD.loc[ASD_DNV_no_seeds].dropna(),color='#0ED50A',label='ASD DNVs',kde=True,hist_kws={'histtype':'stepfilled',\n",
    "                                                                                                     'lw':0},\n",
    "            kde_kws={'lw':.5})\n",
    "plt.xticks(fontsize=8)\n",
    "plt.yticks(fontsize=8)\n",
    "top_DNVs = z_ASD.loc[ASD_DNV_no_seeds].dropna().sort_values(ascending=False)\n",
    "top_DNVs = top_DNVs[top_DNVs>3]\n",
    "#sns.rugplot(z_ASD.loc[top_DNVs.index.tolist()],color='k',height=.1,lw=.2)\n",
    "ax = plt.gca()\n",
    "#ax.scatter([], [], marker=\"|\", linewidth=.5, color='k',s=100, label='ASD DNVs $z_{ASD}>3$') # so legend works\n",
    "ax.set_xlabel('network proximity, $z_{ASD}$',fontsize=8)\n",
    "ax.set_ylabel('density',fontsize=8)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.legend(loc='upper right',fontsize=6,frameon=False,\n",
    "          bbox_to_anchor=(.7,1.35))\n",
    "print(top_DNVs.head())\n",
    "print(ks_2samp(z_ASD.loc[np.setdiff1d(z_ASD.index.tolist(),ASD_HC)],z_ASD.loc[ASD_DNV_no_seeds].dropna()))\n",
    "plt.xlim([-7,6])\n",
    "plt.ylim([0,.43])\n",
    "\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/z_ASD_dist.png',bbox_inches='tight',dpi=300)\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/z_ASD_dist.svg',bbox_inches='tight',dpi=300)\n",
    "\n",
    "\n",
    "# plt.figure(figsize=(8,3))\n",
    "plt.figure(figsize=(2.3,1.54))\n",
    "sns.distplot(z_CHD.loc[np.setdiff1d(z_CHD.index.tolist(),CHD_HC)].dropna(),label='null model',color='#7E7E7E',kde=True,\n",
    "            hist_kws={'histtype':'stepfilled','lw':0},\n",
    "            kde_kws={'lw':.5})\n",
    "CHD_DNV_no_seeds = list(np.setdiff1d(CHD_DNV,CHD_HC))\n",
    "sns.distplot(z_CHD.loc[CHD_DNV_no_seeds].dropna(),color='#C410C4',label='CHD DNVs',kde=True,hist_kws={'histtype':'stepfilled',\n",
    "                                                                                                     'lw':0},\n",
    "            kde_kws={'lw':.5})\n",
    "top_DNVs = z_CHD.loc[CHD_DNV_no_seeds].dropna().sort_values(ascending=False)\n",
    "top_DNVs = top_DNVs[top_DNVs>3]\n",
    "plt.xticks(fontsize=8)\n",
    "plt.yticks(fontsize=8)\n",
    "#sns.rugplot(z_CHD.loc[top_DNVs.index.tolist()],color='k',height=.1,lw=.2)\n",
    "ax = plt.gca()\n",
    "#ax.scatter([], [], marker=\"|\", linewidth=.5, color='k',s=100, label='CHD DNVs $z_{CHD}>3$') # so legend works\n",
    "ax.set_xlabel('network proximity, $z_{CHD}$',fontsize=8)\n",
    "ax.set_ylabel('density',fontsize=8)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.legend(loc='upper right',fontsize=6,frameon=False,\n",
    "          bbox_to_anchor=(.7, 1.35))\n",
    "print(top_DNVs.head())\n",
    "print(ks_2samp(z_CHD.loc[np.setdiff1d(z_CHD.index.tolist(),CHD_HC)],z_CHD.loc[CHD_DNV_no_seeds].dropna()))\n",
    "plt.xlim([-7,6])\n",
    "plt.ylim([0,.43])\n",
    "\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/z_CHD_dist.png',bbox_inches='tight',dpi=300)\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/z_CHD_dist.svg',bbox_inches='tight',dpi=300)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# How many DNVs are found in zscore >= 2? ASD, CHD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def calc_enrichment(z_Gthresh,DNV_list,BG_genes,print_flag=True):\n",
    "    '''\n",
    "    Helper function to calculate hypergeometric stats\n",
    "    '''\n",
    "    #z_Gthresh = z[z>zthresh].index.tolist()\n",
    "    genes_Gthresh = np.intersect1d(DNV_list,z_Gthresh)\n",
    "    \n",
    "    if print_flag==True:\n",
    "        print(len(genes_Gthresh))\n",
    "        print(len(DNV_list))\n",
    "        print(len(z_Gthresh))\n",
    "        print(len(BG_genes))\n",
    "\n",
    "    M = len(BG_genes)\n",
    "    n=len(DNV_list)\n",
    "    N = len(z_Gthresh)\n",
    "    k=len(genes_Gthresh)\n",
    "\n",
    "    p_hyper = scipy.stats.hypergeom.sf(k,M,n,N)\n",
    "        \n",
    "    # calculate expected mean and std given M,n,N\n",
    "    exp_mean = scipy.stats.hypergeom.mean(M,n,N)\n",
    "    exp_std = scipy.stats.hypergeom.std(M,n,N)\n",
    "    \n",
    "    return genes_Gthresh,p_hyper,exp_mean,exp_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "86\n",
      "583\n",
      "1503\n",
      "19655\n",
      "ASD enrichment = 1.1421874132784673e-09\n",
      "62\n",
      "624\n",
      "1000\n",
      "19655\n",
      "CHD enrichment = 1.657817739038242e-07\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'number prioritized genes\\nharboring ASD or CHD dDNVs')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 108x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "zthresh = 2.0\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# single ASD z-scores\n",
    "\n",
    "z_ASD_noseeds = z_ASD.drop(list(np.intersect1d(z_ASD.index,CHD_HC+ASD_HC)))\n",
    "ASD_DNV_in_graph = list(np.intersect1d(ASD_DNV,Gint.nodes()))\n",
    "ASD_DNV_noseeds = list(np.setdiff1d(ASD_DNV_in_graph,ASD_HC))\n",
    "z_ASD_gthresh_noseeds=z_ASD_noseeds[z_ASD_noseeds>=zthresh].index.tolist()\n",
    "ASD_g1,ASD_p,exp_ASD_mean,exp_ASD_std=calc_enrichment(z_ASD_gthresh_noseeds,ASD_DNV_noseeds,z_ASD_noseeds.index.tolist())\n",
    "print('ASD enrichment = ' + str(ASD_p))\n",
    "\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# single CHD z-scores\n",
    "\n",
    "z_CHD_noseeds = z_CHD.drop(list(np.intersect1d(z_CHD.index,CHD_HC+ASD_HC)))\n",
    "CHD_DNV_in_graph = list(np.intersect1d(CHD_DNV,Gint.nodes()))\n",
    "CHD_DNV_noseeds = list(np.setdiff1d(CHD_DNV_in_graph,CHD_HC))\n",
    "z_CHD_gthresh_noseeds=z_CHD_noseeds[z_CHD_noseeds>=zthresh].index.tolist()\n",
    "CHD_g1,CHD_p,exp_CHD_mean,exp_CHD_std=calc_enrichment(z_CHD_gthresh_noseeds,CHD_DNV_noseeds,z_CHD_noseeds.index.tolist())\n",
    "print('CHD enrichment = ' + str(CHD_p))\n",
    "\n",
    "\n",
    "# -------------------------------------------\n",
    "# plot the results\n",
    "# plt.figure(figsize=(3,4))\n",
    "plt.figure(figsize=(1.5,2))\n",
    "\n",
    "recovered_list = [len(ASD_g1),len(CHD_g1)]\n",
    "pval_list = [ASD_p,CHD_p]\n",
    "\n",
    "plt.bar([0,1],recovered_list,width=.4,color='k',label='observed, z>3')\n",
    "plt.bar([0.4,1.4],[exp_ASD_mean,exp_CHD_mean],width=.4,color='gray',\n",
    "       label='expected, z>3')\n",
    "plt.errorbar([0.4,1.4],[exp_ASD_mean,exp_CHD_mean],\n",
    "             [exp_ASD_std,exp_CHD_std],fmt='k.',ecolor='k',capsize=5.0)\n",
    "\n",
    "# add stars for significance\n",
    "for i in [0,1]:\n",
    "    star_height = recovered_list[i]+max(recovered_list)*0.01 # add 5% of max to value\n",
    "    if (pval_list[i] < 0.05) & (pval_list[i]>0.01):\n",
    "        plt.annotate('*',(i,star_height),fontsize=8,ha='center')\n",
    "    elif (pval_list[i] <= 0.01) & (pval_list[i]>0.001):\n",
    "#         plt.annotate('*',[i-.05,i+.05],[star_height]*2,fontsize=6)\n",
    "        plt.annotate('* *',(i,star_height),fontsize=8,ha='center')\n",
    "    elif (pval_list[i] <= 0.001):\n",
    "#         plt.annotate('*',[i-.1,i,i+.1],[star_height]*3,fontsize=6)\n",
    "        plt.annotate('* * *',(i,star_height),fontsize=8,ha='center')\n",
    "                \n",
    "        \n",
    "\n",
    "\n",
    "plt.legend(loc='best',fontsize=6,frameon=False)\n",
    "# plt.legend(loc='upper right',fontsize=6,frameon=False,\n",
    "#           bbox_to_anchor=(1.3, 1.05))\n",
    "plt.xticks([0.2,1.2],['ASD','CHD'],fontsize=8)\n",
    "plt.yticks(fontsize=8)\n",
    "plt.ylabel('number prioritized genes\\nharboring ASD or CHD dDNVs',fontsize=8)\n",
    "# plt.ylim([0,30])\n",
    "\n",
    "# plt.savefig('../../manuscript/figures_1911/Figure2/z_score_num_recovered_DNVs_single_3.0.png',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "# plt.savefig('../../manuscript/figures_1911/Figure2/z_score_num_recovered_DNVs_single_3.0.pdf',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "# plt.savefig('../../manuscript/figures_1911/Figure2/z_score_num_recovered_DNVs_single_3.0.svg',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 1.6)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 108x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# ----------- plot the odds ratios -----------\n",
    "plt.figure(figsize=(1.5,2))\n",
    "recovered_list = [len(ASD_g1),len(CHD_g1)]\n",
    "exp_list_mean = [exp_ASD_mean,exp_CHD_mean]\n",
    "exp_list_std = [exp_ASD_std,exp_CHD_std]\n",
    "obs_div_exp = np.divide(recovered_list,exp_list_mean)\n",
    "lower_ebar=np.divide(recovered_list,np.add(exp_list_mean,exp_list_std))\n",
    "upper_ebar=np.divide(recovered_list,np.subtract(exp_list_mean,exp_list_std))\n",
    "pval_list = [ASD_p,CHD_p]\n",
    "\n",
    "# plt.bar([0,1,2],obs_div_exp,width=.4,color='k',label='observed')\n",
    "# plt.bar([0,1,2],obs_div_exp,color='k',width=.4)\n",
    "log_obs_exp=np.log2(obs_div_exp)\n",
    "log_yerr_lower=np.subtract(np.log2(obs_div_exp),np.log2(lower_ebar))\n",
    "log_yerr_upper=np.subtract(np.log2(upper_ebar),np.log2(obs_div_exp))\n",
    "plt.errorbar([0,1],log_obs_exp,yerr=[log_yerr_lower,\n",
    "                                       log_yerr_upper],fmt='.',color='k',\n",
    "            alpha=.5)\n",
    "plt.plot([0,1],np.log2(obs_div_exp),'k.')\n",
    "ax = plt.gca()\n",
    "# ax.semilogy()\n",
    "plt.ylabel('log2(observed/expected)\\nASD or CHD dDNVs',fontsize=8)\n",
    "plt.xlim([-.5,1.5])\n",
    "plt.xticks([0,1],['ASD\\ngenes','CHD\\ngenes'],fontsize=8,rotation='vertical')\n",
    "plt.plot([-1,3],[0,0],'k--',alpha=.3)\n",
    "\n",
    "# add stars for significance\n",
    "for i in [0,1]:\n",
    "    star_height = log_obs_exp[i]+log_yerr_upper[i]*1.25 # add 25% of max to value\n",
    "    if (pval_list[i] < 0.05) & (pval_list[i]>0.01):\n",
    "        #plt.plot([i],[star_height],'k*')\n",
    "        plt.annotate('*',(i,star_height),fontsize=10,ha='center')\n",
    "    elif (pval_list[i] <= 0.01) & (pval_list[i]>0.001):\n",
    "        #plt.plot([i-.05,i+.05],[star_height]*2,'k*')\n",
    "        plt.annotate('**',(i,star_height),fontsize=10,ha='center')\n",
    "    elif (pval_list[i] <= 0.001):\n",
    "        #plt.plot([i-.1,i,i+.1],[star_height]*3,'k*')\n",
    "        plt.annotate('***',(i,star_height),fontsize=10,ha='center')\n",
    "        \n",
    "plt.ylim([-.5,1.6])\n",
    "# plt.savefig('/Users/brinrosenthal/Google Drive/UCSD/CCBB_tickets/Ideker_17_05_neuro/manuscript/figures_1911/Supplement/obs_div_exp_ind_disease_DNV.png',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "# plt.savefig('/Users/brinrosenthal/Google Drive/UCSD/CCBB_tickets/Ideker_17_05_neuro/manuscript/figures_1911/Supplement/obs_div_exp_ind_disease_DNV.svg',\n",
    "#            dpi=300,bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19\n",
      "192\n",
      "1503\n",
      "19655\n",
      "ASD enrichment = 0.09798447227716468\n",
      "11\n",
      "192\n",
      "1000\n",
      "19655\n",
      "CHD enrichment = 0.2728774966057053\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Text(0,0.5,'number recovered control DNVs')"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 216x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# how many Control DNVs are recovered by ASD, EPI and CHD, respectively\n",
    "\n",
    "\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# control DNVs in single ASD z-scores\n",
    "\n",
    "z_ASD_noseeds = z_ASD.drop(list(np.intersect1d(z_ASD.index,CHD_HC+ASD_HC)))\n",
    "ctrl_DNV_in_graph = list(np.intersect1d(ctrl_DNV,Gint.nodes()))\n",
    "ctrl_DNV_noseeds = ctrl_DNV_in_graph\n",
    "z_ASD_gthresh_noseeds=z_ASD_noseeds[z_ASD_noseeds>=zthresh].index.tolist()\n",
    "ASD_g1,ASD_p,exp_ASD_mean,exp_ASD_std=calc_enrichment(z_ASD_gthresh_noseeds,ctrl_DNV_noseeds,z_ASD_noseeds.index.tolist())\n",
    "print('ASD enrichment = ' + str(ASD_p))\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# control DNVs in single CHD z-scores\n",
    "\n",
    "z_CHD_noseeds = z_CHD.drop(list(np.intersect1d(z_CHD.index,CHD_HC+ASD_HC)))\n",
    "z_CHD_gthresh_noseeds=z_CHD_noseeds[z_CHD_noseeds>=zthresh].index.tolist()\n",
    "CHD_g1,CHD_p,exp_CHD_mean,exp_CHD_std=calc_enrichment(z_CHD_gthresh_noseeds,ctrl_DNV_noseeds,z_CHD_noseeds.index.tolist())\n",
    "print('CHD enrichment = ' + str(CHD_p))\n",
    "\n",
    "# -------------------------------------------\n",
    "# plot the results\n",
    "plt.figure(figsize=(3,4))\n",
    "\n",
    "recovered_list = [len(ASD_g1),len(CHD_g1)]\n",
    "pval_list = [ASD_p,CHD_p]\n",
    "\n",
    "plt.bar([0,1],recovered_list,width=.4,color='k',label='observed')\n",
    "plt.bar([0.4,1.4],[exp_ASD_mean,exp_CHD_mean],width=.4,color='gray',\n",
    "       label='expected')\n",
    "plt.errorbar([0.4,1.4],[exp_ASD_mean,exp_CHD_mean],\n",
    "             [exp_ASD_std,exp_CHD_std],fmt='k.',ecolor='k',capsize=5.0)\n",
    "\n",
    "# add stars for significance\n",
    "for i in [0,1]:\n",
    "    star_height = recovered_list[i]+max(recovered_list)*0.05 # add 5% of max to value\n",
    "    if (pval_list[i] < 0.05) & (pval_list[i]>0.01):\n",
    "        plt.plot([i],[star_height],'k*')\n",
    "    elif (pval_list[i] <= 0.01) & (pval_list[i]>0.001):\n",
    "        plt.plot([i-.05,i+.05],[star_height]*2,'k*')\n",
    "    elif (pval_list[i] <= 0.001):\n",
    "        plt.plot([i-.1,i,i+.1],[star_height]*3,'k*')\n",
    "                \n",
    "\n",
    "\n",
    "plt.legend(loc='best',fontsize=12)\n",
    "plt.xticks([0.2,1.2],['ASD','CHD'],fontsize=16)\n",
    "plt.ylabel('number recovered control DNVs',fontsize=16)\n",
    "\n",
    "\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/z_score_num_recovered_DNVs_single_'+str(zthresh)+'_ctrl.png',\n",
    "#           dpi=300,bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 1.6)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAJEAAACbCAYAAABiQ3lFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAEMxJREFUeJztnX+YlWWZxz/DIDJJDLDBCFxciyLni0qUSpAm/srfK1uybW2mG2iSP7rIvEqkYE0hxkFTcy1bTUHTSkvbVMSt1DUxlTS7FFdvEMModhFF0VHUcZj943kPHN5zZs47nB/vGeb+XNe55pzneee89zV8eZ7nvZ/nvu+6jo4OHKcU+qRtgNPzcRE5JeMickrGReSUjIvIKRkXkVMyLiKnZFxETsm4iJyS6Zu2AeVk8uTJHSNHjkzbjF2GZ5999hUzG1rsul1KRCNHjuTOO+9M24xdBkkvJbkudRFJmgy0mNkRsfbzgTOAjVHTl83Mqmyek4BURSTpAuA04K0C3QcC/2pmT1bXKqe7pL2wXgNM66TvIGCOpOWS5lTRJqebpCoiM7sDaOuk+2fAWcBRwKGSTqqaYU63SH1NVAhJdcBVZrY5+rwUOAC4p8C1M4GZACNGjKimmU5ETYoIGAislLQvYb10FHBjoQvN7DrgOoBp06b5CbsUqCkRSToFGGBm10n6JvAg8C5wv5ndm651TmekLiIzWwt8PHr/k5z2HwM/Tskspxuk/XTm7AK4iJyScRE5JeMickqmy4W1pP7A+cARhKek3wFXmFl75U1zegrFRqK5wD1mdqyZTQV+A/xb5c1yehJdjkRmNjf2+U/AnypqkdPjKDadLQP2BtYD7UCTmX24GoY5PYcupzMzOwFYamZHmtnRwK+rY5bTk0jisW6SND56/6FKGuP0TJKI6ELgbGAr4Od6nDyS+IkmAAOAlYAqa47TE0kiomlAK7AUOLey5jg9kSQiagUwszcJDkfH2YEkIvo9cJykFcB9FbbH6YEkEdFmMzvIzCYBfnLQyaPTpzNJo4D/AMZKWh01twK3VMMwp+fQqYjMbB1woqTDCFsdg9keSOg420gynZ1NcDLuCVxfbgMkTZb03wXap0r6g6RHJZ1Z7vs65SOJiNab2Ytm9jjwSjlvHkXA/gjoH2vfDbgSOBY4HJgpac9y3tspH0k81lskXU7wWG8t8/2zEbDxA/n7Ai+Y2WsAkpYDU4Cfl/n+ThkoKiIzmytpCrDJzJ7N7ZN0FvBb4BrgETOb352bm9kdkkYX6BoIbM75/CbQ2J3vdqpHURFJWgTsBfxK0gwz+3pO90eAV4EbgEPKaNcbwAdzPn8QeL0T+zwCNmWSrIkagNVmdgthDy2XPYBTCPtqe5XRrucIroUhkvoBhwGPFrrQzK4zs4lmNnHw4MFlNMFJSpI1UV9gsKQjgGHZRkl9gW8DQwihzrNLNSYWAXs+8F8Eod9oZn8r9fudypBERN8BvgKcAJyX034fYT30QzMrONUkoYsI2LuBu3f2e53qkWQ6Gxu99ib4igCITjquAK6RdKWkvStjolPrJBmJjgE+A+wGfJ8gHADM7AHgAUnHAQ8BoyphpFPbJBHRe0AdYfO1NdsYrYn+CfgcsAo4uBIGOrVPEhEdBNxFmPp2k3SvmZ1IGHl+CpxmZoVyLjq9hCQiWhBteSBpmpllc/zOJYQRHSSFU7Nm9ruKWOnUNElEdIakQ4G/BzYBWRFNyrnmYML5633La57TE0gioiuBmwkC+kK20cxaACSdR1grfawSBjq1TxIRzQeOBwYBS4CTACQNIBxae9zMTq2UgWmyZMkSAKZPn56qHbVOEj/RTMLUVU/Y4sjyJGHbo0HSBdGxDqcXkmQkugJ4H7gJWAB8NmqfGf3sILgAnF5KkpHodeBlM3sYeDmnfQUhNfAs4EjgqfKb5/QEkojofeCjkr5NOOeT5VpCvqJTgYcJ3mwnBZYsWbJt/ZYGxVLL7A98AziOILjcQ2d9zWx59P5+SadXxkSn1ik2Ek0mrIkmAn+JpdmLH5X1FHy9lGKZ0m4EkDQI+LSk083s/Kj7k5LuZfuiekLlzHRqmWLT2bVsj3qty3mPmXmdTAco/oh/KTAPuJywwM6OQtlUfI3A24SpbPd49USnd1As3d5LQJuZPW9mL5CzDopS8T1OOG90PP6I32tJ4mxcIekegoBui/UNI3it+xDOWncLSX2AHxCiRt4FvhSJNdt/NfAJQsgQwKeyNdCc2iGJiH4FbCAEGv411ncZ8EOCwC7bift/GuhvZgdL+jjwXeBTOf0HAseZWVkjb53y0t1tj/ls3/bI5rUuZfP1UKKcR2b2mKSJ2Y5olBoLXCepCbgh+7To1BalbHuUg3ika3t07BbCNPnvBJEeD5wjKc+NIGmmpCckPfHaa6+V2TwnCUlGooLbHpImxS80sxXxtiLEI137mNn70fu3ge+Z2dvR/R4grJ2ejt3Ty3emTNLpbAL52x4nRD9z/+G6K6JHgKnA7dGa6JmcvgzwM0kHRvc+lDClOjVGkunsYsKJxjp2fMS/GPhPYH9gPHDHTtz/l8A7kn5POEH5NUnnS/pHM3sOuBV4jBAUcHM8oYRTI3R0dBR9ZTKZPTKZzCWZTObPsfZbM5nM4Oj1oyTfVcnXySef3FEuFi5c2NHQ0NABdDQ2NnYsXLiwbN9dbhYvXtyxePHisn9vJpN5oiPB3z1JVpAzCcmmVgFHx7pbc3IIbSm7wlOiubmZefPm0d4e9pQ3b97MvHnzAJgzx4sKxEmyJtoI/EsnhfJWSbqDMM09XlbLUqSlpWWbgLK0t7fT0tLiIipAEhGdACyj8FGPVuA0wlNVa4H+HsnmzYWd4p2193aSLKz3BJZKWhYd/cjlcMJT036FHvl7Ko2NhZOyddbe20kiohmEx/zzyPFWRxghcPEEgkNwl2D27NnU19fv0FZfX8/s2SWnYNol2altjyi762BCcqubCY//F1bKyGqTXffMnz+fLVu20NjYyOzZs3091AlJRPQ68LaZPSzpc1HbIMJZowOB/QgOx1WVMTEd5syZw/DhwwEPXixG0m2Pj+Rue5jZRuB0SYMJjsY+eN2PXksSEV1A8BPVAZfE+q4nlGxoI4jIs4L0QpIsrC8C1gHfIpz3yeVJM1tgZi1mtqjs1jlFaW5u5pxzzmHGjBkMGjSI5ubmqtuQZCQaSFj7XEq+x3qqpHGEBTZmdk55zXO6olY860lGor8jJHJ4iZCjKJfPE040Xgu0lNc0pxhdedarSZKR6KvABwi5G2fG+j5DOAO9lbAeurqs1jldUiue9SQi+hIho/07wIOEg/VZ9jGzaQCS4uslp8I0NjYWFEy1PetJprNRZjbVzP6Z/IiOQZIaJTWSk23fqQ614lkvFgE7GRgu6ZOExfPo2CWXEqI96qL3ThWpFc96senseELxl09En9fF+vcHFgH7AOMIPiOnitSCZ71YBGz8COwvYpcca2ZPmdnP2YU2YJ3ukdRjfVb0/jLCQjvL1igBaB07kXIvQQTsmcCXCVsvC8zsnu7ew6k8SUTU1RHYKwhrot0ocwRsVPN1FiE3Un9guaTfmNm7O3GfncI3XpORRESdHoE1s5VUKAKWkLH2kUg070p6gRC69IfOvqytrQ2ALVu28NRT+fklxowZQ1NTE62trTz99NN5/WPHjmXo0KG88cYbrFy5Mq9/3LhxDBkyhE2bNvH888/n9Y8fP56BAweyceNGVq9endc/YcIEBgwYwIYNG1izZk1e/wEHHEBDQwPr169n7dq1ef0TJ06kX79+rFu3jnXrti9Ps/dqb2+nvr6etWvXsn79+rzfP+SQUBxzzZo1bNiwYYe++vp6Jk+eDMCqVat45ZXkkevFns6Gm9l3ozxFfcysVdIIM8u3cOcoGAEbBTAmqgObW75z2DD3MqRBsZHo85EP6FGgTdLhwHrCFFYOuoqATVQHtlAEbENDw7b/dYUYMGBAl/0DBw7ssn/IkCFd9g8dOpShQ4d22t/U1ERTU1On/SNGjOiynu2oUaMYNWp7VbBVq8JRrqzPaPTo0YwePbrT3x8zZgxjxozptD+TyZDJZDrtj1Ps6ewK4BZC3Y79gSVmVi4BQYiAPRGgQATsCmCKpP6RkPcluBucGqPYdPYxwuL3GjP7X0kzgBfLeP9fAsdEEbB1wIyo9usLZnZXlJ/oYYLYv2Vm75Tx3k6ZKDadzQYWAldIOpeQ9HxxuW5uZlvZ7j7I8nxO//WEg29ODVNs72yTmf2RsJN/eYLrnV5IMVH8VdJnzexl4AbyD6U5TtGF9SXAQ5KGAasJmcscZweSTE+/JcScLQaekhQ/rO/0cpJ4rB80s1kAki4D+lXWJKenkURETZI+TAgJGkEo4+k420giolnAucDuwBz8BKMTI4mIxhOKBNcBe5nZQ5U1yekuaZ82SLKwnk4oHnwSO9aAdRwg2UjUls2SJslrvTp5FNs7mwS8KOm+qGl5V9c7vZNiI1E2V/Wj0U/f9nDyKFZ58eJqGeL0XHxkcUrGReSUjIvIKRkXkVMyLiKnZJI4GyuCpAZCEMAwQjjQF6OEornX3EVIstUGbImKFzs1Rpoj0dnAM2Y2hZALe26Ba/YBDjWzI1xAtUuaItoW/UqoHbLD0duo7usg4G5JyyWdVGX7nIRUZTqTdAbwtVjzBrZHuBaKbu1HiM3/HiG51iOSVkTnvXO/e1sELNAqycppO/AhoCdUw66EnfEcnYVJUhStEq9MJnNnJpOZFL1vzGQyK2P9u2UymT1yPt+eyWSmpGBnosJxab/StDPN6Wxb9Cthj+7hWP/RwO0AUfqa8cBzVbPOSUxqT2eEtMU3SVpOyEx7CoCkRcAvzGyZpOMkPUbISPJNM+sJ00rvI+1huNZfmUxmZto21LqddR0dXtfFKQ33WDsl4yJySsZF5JRMmk9nPYYoyy0EN8O6bCLUWiMtO30kykHSHp103URIRDofuKp6FnWbVOx0EQGSpkj6NXCDpPslHRm7ZAPhH2cRIZdkrZKKnT6dBWYB/2BmbZL6AbcRKipl6QccQ4h6aUjBPmDbpnQz4fjMRmCumf0t55JU7PSRKPCOmbUBmNl7RJUkc/g6oYzpeuDMKtuWy1XAfDM7CfgO+VNWKnb6SBTYPUrkBSHnQDx9ztXAHoS9vPnA6VW0LZf3zezPAGb2gqR4dYFU7PSRKPAWofxoC6HkVnwkepfwtHM3EC9NUU3i2wtbY59TsdNHIsDMZkRrofeBg4EnY5fUA6MlTQc6z2JeeSZIupftxXiGx/pTsdNFBEi6mJCIPUNYkH6RHevdzgG+EPXNqLqBEWb20SKXpGKniygwwswuknS1mc2SdFOs/1RCUvitwFS2n4OqKpKWEaa0bacYzSzXllTsdBFFRI7GN6OPH4h1jwOON7NUjzxkgxUkLTazQiNNKna6iAIvEkpEfEXShUBrrH8zYa3xFkD8nHcKdCaSVOx0EQFm1gw0S9oPOIp8R90oYB5hQdtBSo/4OW6Ihuz7mFBSsdNFBEg6mOCo6wu8Z2bHxi5ZABxBqHL0l+patwMthGDOVwlRMO+w4wI6FTvdTxS4A1gKTKPwH/8i4ADgbULN2rTYANwWrYf+SPBM55KKnS6iwP6E/ahHgUmS4hXvNgFvmNkzwP9V27gc9jGzWwHM7DLy48JSsdNFBJjZa2Z2KXAIcCVRqFIOrwKHSVpCmELSYnPsczwRayp2+kH9BEj6H2BvwjTxMvADM7s6BTsWEWLvHiLs1k8ws3PTttNHomQ8YGb9zWwIYe00qtgvVIg5BB/WN6LPX431p2KnP50lY5ikbFHjPUkpNj/KJ/79Li5JxU4XUTIuAq4h1DdpId1N2K5IxU5fEzkl42sip2RcRE7JuIicknEROSXjInJK5v8BPxmB25bzspEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 108x144 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "# ----------- plot the odds ratios -----------\n",
    "plt.figure(figsize=(1.5,2))\n",
    "recovered_list = [len(ASD_g1),len(CHD_g1)]\n",
    "exp_list_mean = [exp_ASD_mean,exp_CHD_mean]\n",
    "exp_list_std = [exp_ASD_std,exp_CHD_std]\n",
    "obs_div_exp = np.divide(recovered_list,exp_list_mean)\n",
    "lower_ebar=np.divide(recovered_list,np.add(exp_list_mean,exp_list_std))\n",
    "upper_ebar=np.divide(recovered_list,np.subtract(exp_list_mean,exp_list_std))\n",
    "pval_list = [ASD_p,CHD_p]\n",
    "\n",
    "# plt.bar([0,1,2],obs_div_exp,width=.4,color='k',label='observed')\n",
    "# plt.bar([0,1,2],obs_div_exp,color='k',width=.4)\n",
    "log_obs_exp=np.log2(obs_div_exp)\n",
    "log_yerr_lower=np.subtract(np.log2(obs_div_exp),np.log2(lower_ebar))\n",
    "log_yerr_upper=np.subtract(np.log2(upper_ebar),np.log2(obs_div_exp))\n",
    "plt.errorbar([0,1],log_obs_exp,yerr=[log_yerr_lower,\n",
    "                                       log_yerr_upper],fmt='o',color='k',\n",
    "            alpha=.5)\n",
    "plt.plot([0,1],np.log2(obs_div_exp),'ko')\n",
    "ax = plt.gca()\n",
    "# ax.semilogy()\n",
    "plt.ylabel('log2(observed/expected)\\ncontrol dDNVs',fontsize=8)\n",
    "plt.xlim([-.5,1.5])\n",
    "plt.xticks([0,1],['ASD\\ngenes','CHD\\ngenes'],fontsize=8,rotation='vertical')\n",
    "plt.plot([-1,3],[0,0],'k--',alpha=.3)\n",
    "\n",
    "# add stars for significance\n",
    "for i in [0,1]:\n",
    "    star_height = log_obs_exp[i]+log_yerr_upper[i]*1.25 # add 25% of max to value\n",
    "    if (pval_list[i] < 0.05) & (pval_list[i]>0.01):\n",
    "        #plt.plot([i],[star_height],'k*')\n",
    "        plt.annotate('*',(i,star_height),fontsize=10,ha='center')\n",
    "    elif (pval_list[i] <= 0.01) & (pval_list[i]>0.001):\n",
    "        #plt.plot([i-.05,i+.05],[star_height]*2,'k*')\n",
    "        plt.annotate('**',(i,star_height),fontsize=10,ha='center')\n",
    "    elif (pval_list[i] <= 0.001):\n",
    "        #plt.plot([i-.1,i,i+.1],[star_height]*3,'k*')\n",
    "        plt.annotate('***',(i,star_height),fontsize=10,ha='center')\n",
    "        \n",
    "plt.ylim([-.5,1.6])\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/obs_div_exp_CTRL_DNV.png',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/obs_div_exp_CTRL_DNV.svg',\n",
    "#            dpi=300,bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "z_ASD.sort_values(ascending=False,inplace=True)\n",
    "z_CHD.sort_values(ascending=False,inplace=True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max ASD_HC zscore:\n",
      "51.9050361869\n",
      "\n",
      "max CHD_HC zscore:\n",
      "34.857337174899996\n"
     ]
    }
   ],
   "source": [
    "print('max ASD_HC zscore:')\n",
    "print(z_ASD.loc[ASD_HC].max())\n",
    "print('\\nmax CHD_HC zscore:')\n",
    "print(z_CHD.loc[CHD_HC].max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plot the shared-phenotype DNVs in combined z-score space\n",
    "\n",
    "## CHD-ASD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "398\n"
     ]
    }
   ],
   "source": [
    "# Load the high confidence genes called by Jeremy (bonferroni p < 0.1, zASD>1.5, zCHD>1.5, zASD-CHD>3)\n",
    "ASD_CHD_high_conf = pd.read_excel('data/supplemental_tables_cell_systems_210416.xlsx',sheet_name='Table S4',skiprows=1)\n",
    "ASD_CHD_high_conf.index=ASD_CHD_high_conf['gene']\n",
    "ASD_CHD_high_conf_genes = ASD_CHD_high_conf[(ASD_CHD_high_conf['corrected pvalue for both (bonferroni)']<.1)&(ASD_CHD_high_conf['lowest Z score']>1.5)].index.tolist()\n",
    "print(len(ASD_CHD_high_conf_genes))\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "combine_after_zscore=True # whether to use the score combined before or after z-score calculation (this is obsolete.. should always be true now)\n",
    "include_seeds=False # plot with or without seeds\n",
    "\n",
    "z_ASD_abs = z_ASD.copy(deep=True)\n",
    "ASD_HC_in_G = list(np.intersect1d(ASD_HC,Gint.nodes()))\n",
    "\n",
    "\n",
    "z_CHD_abs = z_CHD.copy(deep=True)\n",
    "CHD_HC_in_G = list(np.intersect1d(CHD_HC,Gint.nodes()))\n",
    "\n",
    "\n",
    "\n",
    "if include_seeds==False: # remove seeds if false\n",
    "    z_ASD_abs.loc[ASD_HC_in_G]=np.nan # remove the seeds\n",
    "    z_CHD_abs.loc[CHD_HC_in_G]=np.nan # remove the seeds\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13.65848615693\n",
      "14.22708069593\n"
     ]
    }
   ],
   "source": [
    "# use diff btwn max and min to set axes labels\n",
    "print(z_ASD_abs.max()-z_ASD_abs.min())\n",
    "\n",
    "print(z_CHD_abs.max()-z_CHD_abs.min())\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-4.34949947011\n",
      "-4.70643282787\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/brinrosenthal/anaconda3/envs/ipykernel_py2/lib/python2.7/site-packages/ipykernel_launcher.py:77: FutureWarning: \n",
      "Passing list-likes to .loc or [] with any missing label will raise\n",
      "KeyError in the future, you can use .reindex() as an alternative.\n",
      "\n",
      "See the documentation here:\n",
      "https://pandas.pydata.org/pandas-docs/stable/indexing.html#deprecate-loc-reindex-listlike\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/Users/brinrosenthal/Google Drive/UCSD/CCBB_tickets/Ideker_17_05_neuro/manuscript/figures_1911/Figure2/Figure2_final assets/ASD_CHD_zscore_dots_noDNV_no_seeds_combine_after_zscore\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"ticks\", {\"xtick.major.size\": 25, \"ytick.major.size\": 25})\n",
    "\n",
    "# -------------------- Plot on the paired disorder axes ----------------------\n",
    "\n",
    "print(z_ASD.min())\n",
    "print(z_CHD.min())\n",
    "D1name='ASD'\n",
    "D2name='CHD'\n",
    "\n",
    "genes_D1 = ASD_HC\n",
    "genes_D2 = CHD_HC\n",
    "\n",
    "idx_temp = z_CHD_abs.index.tolist()\n",
    "\n",
    "z_ASD_abs = z_ASD_abs.loc[idx_temp]\n",
    "\n",
    "# j = sns.jointplot(z_CHD_abs,z_ASD_abs,gridsize=35,kind='hex',height=10)\n",
    "j = sns.jointplot(z_CHD_abs,z_ASD_abs,kind='scatter',height=8,alpha=.1,joint_kws={'s':6,'color':'gray'},marginal_kws={'color':'white'})\n",
    "# set the colors of the marginal distributions to match other figs\n",
    "# plt.sca(j.ax_marg_x)\n",
    "# sns.distplot(z_CHD_abs.dropna().tolist(),color='#C410C4',kde=False)\n",
    "# plt.sca(j.ax_marg_y)\n",
    "# sns.distplot(z_ASD_abs.dropna().tolist(),color='#0ED50A',kde=False,vertical=True)\n",
    "\n",
    "# plt.ylim([0,.4])\n",
    "\n",
    "plt.sca(j.ax_joint)\n",
    "j.ax_joint.set_xlabel('$z_{CHD}$',fontsize=24)\n",
    "j.ax_joint.set_ylabel('$z_{ASD}$',fontsize=24)\n",
    "\n",
    "zthresh=3\n",
    "\n",
    "if combine_after_zscore == True:\n",
    "    z_score_both = z_CHD_abs*z_ASD_abs*(z_CHD_abs>0)*(z_ASD_abs>0)\n",
    "else:\n",
    "    z_score_both = z_ASD_CHD \n",
    "\n",
    "high_z_genes = z_score_both[z_score_both>=zthresh].index.tolist()\n",
    "plt.plot((z_CHD_abs[z_score_both>=zthresh]),(z_ASD_abs[z_score_both>=zthresh]),'.',color='#F9825C',\n",
    "         ms=4,alpha=.7)\n",
    "\n",
    "# add the high confidence genes\n",
    "# HC_z_genes = z_ASD_abs[(z_ASD_abs>1.5)&(z_CHD_abs>1.5)&(z_score_both>3)].index.tolist()\n",
    "plt.plot((z_CHD_abs.loc[ASD_CHD_high_conf_genes]),(z_ASD_abs.loc[ASD_CHD_high_conf_genes]),'.',color='red',\n",
    "         ms=4,alpha=.5)\n",
    "\n",
    "\n",
    "only_D2=np.setdiff1d(genes_D2,genes_D1)\n",
    "only_D1 = np.setdiff1d(genes_D1,genes_D2)\n",
    "D1_and_D2 = np.intersect1d(genes_D1,genes_D2)\n",
    "\n",
    "# draw the seed nodes with borders\n",
    "only_D2_blue = list(np.intersect1d(only_D2,z_score_both[z_score_both<zthresh].index.tolist()))\n",
    "only_D2_orange = list(np.intersect1d(only_D2,z_score_both[z_score_both>=zthresh].index.tolist()))\n",
    "\n",
    "only_D1_blue = list(np.intersect1d(only_D1,z_score_both[z_score_both<zthresh].index.tolist()))\n",
    "only_D1_orange = list(np.intersect1d(only_D1,z_score_both[z_score_both>=zthresh].index.tolist()))\n",
    "\n",
    "plt.plot((z_CHD_abs.loc[only_D2_orange]),(z_ASD_abs.loc[only_D2_orange]),'o',ms=6,mec='#C815F0',mew=.7,\n",
    "        fillstyle='full',mfc='#F9825C')\n",
    "plt.plot((z_CHD_abs.loc[only_D2_blue]),(z_ASD_abs.loc[only_D2_blue]),'o',ms=6,mec='#C815F0',mew=.7,\n",
    "        fillstyle='full',mfc='gray')\n",
    "plt.plot((z_CHD_abs.loc[only_D1_blue]),(z_ASD_abs.loc[only_D1_blue]),'o',ms=6,mec='#55CCCA',mew=.7,\n",
    "        fillstyle='full',mfc='gray')\n",
    "plt.plot((z_CHD_abs.loc[only_D1_orange]),(z_ASD_abs.loc[only_D1_orange]),'o',ms=6,mec='#55CCCA',mew=.7,\n",
    "        fillstyle='full',mfc='#F9825C')\n",
    "plt.plot((z_CHD_abs.loc[D1_and_D2]),(z_ASD_abs.loc[D1_and_D2]),'o',ms=6,mec='black',mew=.7,\n",
    "        fillstyle='full',mfc='#F9825C')\n",
    "\n",
    "\n",
    "plt.tick_params(length=12)\n",
    "plt.tick_params(width=2)\n",
    "# overlay the DNVs with NDDs\n",
    "# plt.plot(z_CHD_abs.loc[DNV_withNDD],z_ASD_abs.loc[DNV_withNDD],'k^',ms=4.5)\n",
    "\n",
    "#annotate top DNVs -- add 10 so we don't have to deal with multiplying negative numbers\n",
    "DNVs_top_10 = ((z_ASD_abs.loc[DNV_withNDD]+10)*(z_CHD_abs.loc[DNV_withNDD]+10)).dropna().sort_values(ascending=False).head(16).index.tolist()\n",
    "# if include_seeds==False: # only print labels if not including seeds\n",
    "#     for d in DNVs_top_10:\n",
    "#             print(d)\n",
    "# #             plt.annotate(d,xy=(z_CHD_abs.loc[d]-.03,z_ASD_abs.loc[d]-.01),\n",
    "# #                        fontsize=9,rotation=0,horizontalalignment='right')\n",
    "#     texts_tmp = [plt.text(z_CHD_abs.loc[d],z_ASD_abs.loc[d],d) for d in DNVs_top_10]\n",
    "#     adjustText.adjust_text(texts_tmp,avoid_points=False,expand_text=(1.05,1.05),lim=10)\n",
    "\n",
    "ax=plt.gca()\n",
    "\n",
    "if include_seeds==False: # only set axes limits if not including seeds\n",
    "    plt.xlim([-4,6])\n",
    "    plt.ylim([-4,6])\n",
    "plt.xticks(fontsize=20)\n",
    "plt.yticks(fontsize=20)\n",
    "\n",
    "plt.plot([-5,7],[0,0],color='k',lw=1)\n",
    "plt.plot([0,0],[-5,7],color='k',lw=1)\n",
    "#ax.set_yscale('log')\n",
    "#ax.set_xscale('log')\n",
    "\n",
    "# fname_string = '../../manuscript/figures_1911/'+focal_interactome+'/ASD_CHD_zscore_dots_'\n",
    "# ------------------------ if combined post-propagation ------------------------\n",
    "fname_string = '/Users/brinrosenthal/Google Drive/UCSD/CCBB_tickets/Ideker_17_05_neuro/manuscript/figures_1911/Figure2/Figure2_final assets/ASD_CHD_zscore_dots_noDNV'\n",
    "# ------------------------ if combined pre-propagation ------------------------\n",
    "# fname_string = '../../manuscript/figures_1911/Supplement/ASD_CHD_zscore_dots_noDNV'\n",
    "if include_seeds==True:\n",
    "    fname_string+='_with_seeds'\n",
    "    \n",
    "elif include_seeds==False:\n",
    "    fname_string+='_no_seeds'\n",
    "    \n",
    "if combine_after_zscore==True:\n",
    "    fname_string+='_combine_after_zscore'\n",
    "elif combine_after_zscore==False:\n",
    "    fname_string+='_combine_before_zscore'\n",
    "    \n",
    "print(fname_string)\n",
    "# plt.savefig(fname_string+'_201012.png',dpi=300,bbox_inches='tight')\n",
    "# plt.savefig(fname_string+'_201012.svg',dpi=300,bbox_inches='tight')\n",
    "# plt.savefig(fname_string+'.pdf',dpi=300,bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "764"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# number high scoring genes\n",
    "len(high_z_genes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Calculate how many CHD-NDD DNVs are recovered above z = zthresh, in z_ASD, z_CHD, and z_ASD_CHD\n",
    "\n",
    "And compare to number expected given hypergeometric distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7\n",
      "155\n",
      "412\n",
      "19655\n",
      "ASD enrichment = 0.016663258298861512\n",
      "3\n",
      "155\n",
      "242\n",
      "19655\n",
      "CHD enrichment = 0.1248217683401912\n",
      "15\n",
      "155\n",
      "764\n",
      "19655\n",
      "ASD-CHD enrichment = 0.00036114001181785255\n",
      "11\n",
      "155\n",
      "351\n",
      "19655\n",
      "ASD-CHD HC enrichment = 2.3006007484662232e-05\n"
     ]
    }
   ],
   "source": [
    "sns.set_style(\"ticks\", {\"xtick.major.size\": 15, \"ytick.major.size\": 15})\n",
    "\n",
    "\n",
    "zthresh1 = 3 # combined zthreshold\n",
    "\n",
    "DNV_noseeds = list(np.setdiff1d(DNV_withNDD,CHD_HC+ASD_HC))\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# shared phenotype DNVs enriched in single ASD z-scores?\n",
    "    \n",
    "z_ASD_noseeds = z_ASD.drop(list(np.intersect1d(z_ASD.index,CHD_HC+ASD_HC)))\n",
    "z_ASD_gthresh_noseeds=z_ASD_noseeds[z_ASD_noseeds>=zthresh1].index.tolist()\n",
    "ASD_g1,ASD_p1,exp_ASD1_mean,exp_ASD1_std=calc_enrichment(z_ASD_gthresh_noseeds,DNV_noseeds,z_ASD_noseeds.index.tolist())\n",
    "print('ASD enrichment = ' + str(ASD_p1))\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# shared phenotype DNVs enriched in single CHD z-scores?\n",
    "\n",
    "    \n",
    "z_CHD_noseeds = z_CHD.drop(list(np.intersect1d(z_CHD.index,CHD_HC+ASD_HC)))\n",
    "z_CHD_gthresh_noseeds=z_CHD_noseeds[z_CHD_noseeds>=zthresh1].index.tolist()\n",
    "CHD_g1,CHD_p1,exp_CHD1_mean,exp_CHD1_std=calc_enrichment(z_CHD_gthresh_noseeds,DNV_noseeds,z_CHD_noseeds.index.tolist())\n",
    "print('CHD enrichment = ' + str(CHD_p1))\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# shared phenotype DNVs enriched in combined ASD-CHD z-scores?\n",
    "\n",
    "z_both = z_CHD*z_ASD*(z_CHD>0)*(z_ASD>0)\n",
    "z_both_noseeds = z_both.drop(list(np.intersect1d(z_CHD.index,CHD_HC+ASD_HC)))\n",
    "z_both_gthresh_noseeds=z_both_noseeds[z_both_noseeds>=zthresh1].index.tolist()\n",
    "both_g1,both_p1,exp_both1_mean,exp_both1_std=calc_enrichment(z_both_gthresh_noseeds,DNV_noseeds,z_both_noseeds.index.tolist())\n",
    "print('ASD-CHD enrichment = ' + str(both_p1))\n",
    "\n",
    "# ---------------------------------------------------------------------------\n",
    "# shared phenotype DNVs enriched in combined ASD-CHD high confidence genes (ASD>1.5 and CHD>1.5, fdr<0.1)?\n",
    "\n",
    "\n",
    "ASD_CHD_high_conf_no_seeds=list(np.setdiff1d(ASD_CHD_high_conf_genes,ASD_HC+CHD_HC))\n",
    "both_HC_g1,both_HC_p1,exp_both_HC1_mean,exp_both_HC1_std=calc_enrichment(ASD_CHD_high_conf_no_seeds,DNV_noseeds,z_ASD_noseeds.index.tolist())\n",
    "print('ASD-CHD HC enrichment = ' + str(both_HC_p1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "929\n",
      "5\n",
      "155\n",
      "929\n",
      "19655\n",
      "ASD specific enrichment = 0.7465660145939729\n",
      "616\n",
      "6\n",
      "155\n",
      "616\n",
      "19655\n",
      "CHD specific enrichment = 0.21405760863520584\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "\n",
    "z_ASD_noseeds = z_ASD.drop(list(np.intersect1d(z_ASD.index,CHD_HC+ASD_HC)))\n",
    "z_ASD_specific=z_ASD_noseeds[(z_both_noseeds<3)&(z_ASD_noseeds>=2)&(z_CHD_noseeds<2)].index.tolist()\n",
    "print(len(z_ASD_specific))\n",
    "ASDspecific_g1,ASDspecific_p1,exp_ASDspecific_mean,exp_ASDspecific_std=calc_enrichment(z_ASD_specific,DNV_noseeds,z_ASD_noseeds.index.tolist())\n",
    "print('ASD specific enrichment = ' + str(ASDspecific_p1))\n",
    "\n",
    "\n",
    "z_CHD_noseeds = z_CHD.drop(list(np.intersect1d(z_CHD.index,CHD_HC+ASD_HC)))\n",
    "z_CHD_specific=z_CHD_noseeds[(z_both_noseeds<3)&(z_CHD_noseeds>=2)&(z_ASD_noseeds<2)].index.tolist()\n",
    "print(len(z_CHD_specific))\n",
    "CHDspecific_g1,CHDspecific_p1,exp_CHDspecific_mean,exp_CHDspecific_std=calc_enrichment(z_CHD_specific,DNV_noseeds,z_CHD_noseeds.index.tolist())\n",
    "print('CHD specific enrichment = ' + str(CHDspecific_p1))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-1.5, 4.1)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 144x144 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"ticks\", {\"xtick.major.size\": 15, \"ytick.major.size\": 15})\n",
    "\n",
    "# ----------- plot the odds ratios -----------\n",
    "plt.figure(figsize=(2,2))\n",
    "recovered_list = [len(ASDspecific_g1),len(CHDspecific_g1),len(both_g1),len(both_HC_g1)]\n",
    "exp_list_mean1 = [exp_ASDspecific_mean,exp_CHDspecific_mean,exp_both1_mean,exp_both_HC1_mean]\n",
    "exp_list_std1 = [exp_ASDspecific_std,exp_CHDspecific_std,exp_both1_std,exp_both_HC1_std]\n",
    "obs_div_exp = np.divide(recovered_list,exp_list_mean1)\n",
    "lower_ebar=np.divide(recovered_list,np.add(exp_list_mean1,exp_list_std1))\n",
    "upper_ebar=np.divide(recovered_list,np.subtract(exp_list_mean1,exp_list_std1))\n",
    "pval_list = [ASDspecific_p1,CHDspecific_p1,both_p1,both_HC_p1]\n",
    "\n",
    "# plt.bar([0,1,2],obs_div_exp,width=.4,color='k',label='observed')\n",
    "# plt.bar([0,1,2],obs_div_exp,color='k',width=.4)\n",
    "log_obs_exp=np.log2(obs_div_exp)\n",
    "log_yerr_lower=np.subtract(np.log2(obs_div_exp),np.log2(lower_ebar))\n",
    "log_yerr_upper=np.subtract(np.log2(upper_ebar),np.log2(obs_div_exp))\n",
    "plt.errorbar([0,1,2,3],log_obs_exp,yerr=[log_yerr_lower,\n",
    "                                       log_yerr_upper],fmt='o',color='k',\n",
    "            alpha=.5)\n",
    "plt.plot([0,1,2,3],np.log2(obs_div_exp),'ko')\n",
    "ax = plt.gca()\n",
    "# ax.semilogy()\n",
    "plt.ylabel('log2(observed/expected)\\ndual-phenotype dDNVs',fontsize=10)\n",
    "plt.xlim([-.5,3.5])\n",
    "plt.yticks([-1,0,1,2,3,4],fontsize=8)\n",
    "plt.xticks([0,1,2,3],['ASD-specific\\nnetwork genes','CHD-specific\\nnetwork genes','ASD-CHD network\\nintersection genes',\n",
    "                   'HC ASD-CHD network\\nintersection genes'],fontsize=10,\n",
    "           rotation='vertical')\n",
    "plt.plot([-1,4],[0,0],'k--',alpha=.3,lw=1)\n",
    "\n",
    "# add stars for significance\n",
    "for i in [0,1,2,3]:\n",
    "    star_height = log_obs_exp[i]+log_yerr_upper[i]*1.25 # add 25% of max to value\n",
    "    if (pval_list[i] < 0.05) & (pval_list[i]>0.01):\n",
    "        #plt.plot([i],[star_height],'k*')\n",
    "        plt.annotate('*',(i,star_height),fontsize=10,ha='center')\n",
    "    elif (pval_list[i] <= 0.01) & (pval_list[i]>0.001):\n",
    "        #plt.plot([i-.05,i+.05],[star_height]*2,'k*')\n",
    "        plt.annotate('**',(i,star_height),fontsize=10,ha='center')\n",
    "    elif (pval_list[i] <= 0.001):\n",
    "        #plt.plot([i-.1,i,i+.1],[star_height]*3,'k*')\n",
    "        plt.annotate('***',(i,star_height),fontsize=10,ha='center')\n",
    "        \n",
    "plt.ylim([-1.5,4.1])\n",
    "# ------------------------ if combined post-propagation ------------------------\n",
    "# plt.savefig('/Users/brinrosenthal/Google Drive/UCSD/CCBB_tickets/Ideker_17_05_neuro/manuscript/figures_1911/Figure2/Figure2_final assets/obs_div_exp_dual_phenotype_dnv_201012.png',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "# plt.savefig('/Users/brinrosenthal/Google Drive/UCSD/CCBB_tickets/Ideker_17_05_neuro/manuscript/figures_1911/Figure2/Figure2_final assets/obs_div_exp_dual_phenotype_dnv_201012.svg',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "\n",
    "\n",
    "# ------------------------ if combined pre-propagation ------------------------\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/obs_div_exp_dual_pheno_combine_before_prop.png',\n",
    "#            dpi=300,bbox_inches='tight')\n",
    "# plt.savefig('../../manuscript/figures_1911/Supplement/obs_div_exp_dual_pheno_combine_before_prop.svg',\n",
    "#            dpi=300,bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.17"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}