Script to Create Correlation table and Heatmap of Drug Expression from RNA-Expression file from GEO or Studies.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b80a72ed",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/cd/1pn_n69n5vzf7yz6_fmnrkjr0000gp/T/ipykernel_89168/3738193197.py:79: FutureWarning: The default value of numeric_only in DataFrameGroupBy.median is deprecated. In a future version, numeric_only will default to False. Either specify numeric_only or select only columns which should be valid for the function.\n",
      "  placebo = placebo.groupby('Gene.name').median().reset_index()\n",
      "/var/folders/cd/1pn_n69n5vzf7yz6_fmnrkjr0000gp/T/ipykernel_89168/3738193197.py:87: FutureWarning: The default value of numeric_only in DataFrameGroupBy.median is deprecated. In a future version, numeric_only will default to False. Either specify numeric_only or select only columns which should be valid for the function.\n",
      "  abmrA = abmrA.groupby('Gene Name').median().reset_index()\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from scipy.stats import pearsonr\n",
    "\n",
    "# Curating expression datasets for studies/GEO files\n",
    "\n",
    "study_1 = pd.read_excel(\"path to file with gene expression from GEO or study\")\n",
    "study_1 = study_1[study_1[\"pvalue\"] <0.05]\n",
    "study_1 = study_1.dropna()\n",
    "\n",
    "study_1 = study_1[['Gene.symbol','adj.P.Val', 'logFC']]\n",
    "study_1 = study_1.rename(columns={\"Gene.symbol\":\"Gene Name\", \"adj.P.Val\":\"pvalue\", \"logFC\":\"logchange\"})\n",
    "\n",
    "listofstudygenes = study_1[\"Gene Name\"].to_list()\n",
    "new_genes = []\n",
    "\n",
    "for i in listofstudygenes:\n",
    "    x = i.split(\"/\")[0]\n",
    "    new_genes.append(x)\n",
    "\n",
    "study_1[\"Gene Name\"] = new_genes\n",
    "study_1 = study_1.groupby('Gene Name').median().reset_index()\n",
    "\n",
    "\n",
    "# Curating expression datasets for drugs\n",
    "\n",
    "p450 = pd.read_csv(\"path to p450 file\")\n",
    "p450 = p450[p450[\"padj\"] <0.05]\n",
    "\n",
    "p900 = pd.read_csv(\"path to p900 file\")\n",
    "p900 = p900[p900[\"padj\"] <0.05]\n",
    "\n",
    "sf6222 = pd.read_csv(\"path to sf6222 file\")\n",
    "sf6222 = sf6222[sf6222[\"padj\"] <0.05]\n",
    "sf6222 = sf6222[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "sf7009 = pd.read_csv(\"path to sf7009 file\")\n",
    "sf7009 = sf7009[sf7009[\"padj\"] <0.05]\n",
    "sf7009 = sf7009[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "sf7044 = pd.read_csv(\"path to sf7044 file\")\n",
    "sf7044 = sf7044[sf7044[\"padj\"] <0.05]\n",
    "sf7044 = sf7044[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "um101 = pd.read_csv(\"path to um101 file\")\n",
    "um101 = um101[um101[\"padj\"] <0.05]\n",
    "um101 = um101[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "tnf = pd.read_csv(\"path to tnf file\")\n",
    "tnf = tnf[tnf[\"padj\"] <0.05]\n",
    "tnf = tnf[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "sf3030 = pd.read_csv(\"path to sf3030 file\")\n",
    "sf3030 = sf3030[sf3030[\"padj\"] <0.05]\n",
    "sf3030 = sf3030[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "pdgf = pd.read_csv(\"path to PDGF file\")\n",
    "pdgf = pdgf[pdgf[\"padj\"] <0.05]\n",
    "\n",
    "bvd = pd.read_csv(\"path to BVD file\")\n",
    "bvd = bvd[bvd[\"padj\"] <0.05]\n",
    "\n",
    "p450_median = p450.groupby('Gene.name').median().reset_index()\n",
    "p900_median = p900.groupby('Gene.name').median().reset_index()\n",
    "sf7009_median = sf7009.groupby('Gene.name').median().reset_index()\n",
    "sf7044_median = sf7044.groupby('Gene.name').median().reset_index()\n",
    "sf6222_median = sf6222.groupby('Gene.name').median().reset_index()\n",
    "sf3030_median = sf3030.groupby('Gene.name').median().reset_index()\n",
    "tnf_median = tnf.groupby('Gene.name').median().reset_index()\n",
    "um101_median = um101.groupby('Gene.name').median().reset_index()\n",
    "pdgf_median = pdgf.groupby('Gene.name').median().reset_index()\n",
    "bvd_median = bvd.groupby('Gene.name').median().reset_index()\n",
    "\n",
    "merged_table = pd.merge(p450_median, p900_median, on=\"Gene.name\", how=\"inner\")\n",
    "p450_900_450 = merged_table[['Gene.name', 'padj_x', \"log2FoldChange_x\"]]\n",
    "p450_900_900 = merged_table[['Gene.name', 'padj_y', \"log2FoldChange_y\"]]\n",
    "\n",
    "merged_table['log2FoldChange'] = (merged_table['log2FoldChange_x'] + merged_table['log2FoldChange_y']) / 2\n",
    "average_450_900 = merged_table[['Gene.name','log2FoldChange']]\n",
    "\n",
    "placebo = pd.read_csv(\"path to placebo file\")\n",
    "placebo = placebo.rename(columns={\"p-value\":\"padj\", \"protein\":\"Gene.name\"})\n",
    "placebo = placebo[placebo[\"padj\"] <0.05]\n",
    "placebo = placebo.groupby('Gene.name').median().reset_index()\n",
    "\n",
    "placebo_inflammation = pd.read_csv(\"/Users/rishabhgoel/Desktop/GEn1E Strategy/Drug Data/Placebo_inflammation.csv\")\n",
    "placebo_inflammation = placebo_inflammation.rename(columns={\"p-value\":\"padj\", \"protein\":\"Gene.name\"})\n",
    "placebo_inflammation = placebo_inflammation[placebo_inflammation[\"padj\"] <0.05]\n",
    "placebo_inflammation = placebo_inflammation.groupby('Gene.name').median().reset_index()\n",
    "\n",
    "\n",
    "#  Option to keep only significant genes with a log fold change threshold\n",
    "\n",
    "# study_1 = study_1[(study_1[\"logchange\"]>1) | (study_1[\"logchange\"]<(-1))]\n",
    "\n",
    "\n",
    "\n",
    "# Naming final instance of curated GEO dataset\n",
    "\n",
    "corr_check = study_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d6bfb97f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Merging all drug datasets with expression dataset\n",
    "\n",
    "average_merge_450_900 = corr_check.merge(average_450_900, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_450 = corr_check.merge(p450_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_900 = corr_check.merge(p900_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_sf7044 = corr_check.merge(sf7044_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_sf6222 = corr_check.merge(sf6222_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_sf7009 = corr_check.merge(sf7009_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_um101 = corr_check.merge(um101_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_sf3030 = corr_check.merge(sf3030_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_pdgf = corr_check.merge(pdgf_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_bvd = corr_check.merge(bvd_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "merge_tnf = corr_check.merge(tnf_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4d6f805a",
   "metadata": {},
   "outputs": [],
   "source": [
    " def create_table(disease_column, corr_check, symbol_col):\n",
    "\n",
    "    p450 = pd.read_csv(\"path to p450 file\")\n",
    "    p450 = p450[p450[\"padj\"] <0.05]\n",
    "\n",
    "    p900 = pd.read_csv(\"path to p900 file\")\n",
    "    p900 = p900[p900[\"padj\"] <0.05]\n",
    "\n",
    "    sf6222 = pd.read_csv(\"path to sf6222 file\")\n",
    "    sf6222 = sf6222[sf6222[\"padj\"] <0.05]\n",
    "    sf6222 = sf6222[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "    sf7009 = pd.read_csv(\"path to sf7009 file\")\n",
    "    sf7009 = sf7009[sf7009[\"padj\"] <0.05]\n",
    "    sf7009 = sf7009[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "    sf7044 = pd.read_csv(\"path to sf7044 file\")\n",
    "    sf7044 = sf7044[sf7044[\"padj\"] <0.05]\n",
    "    sf7044 = sf7044[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "    um101 = pd.read_csv(\"path to um101 file\")\n",
    "    um101 = um101[um101[\"padj\"] <0.05]\n",
    "    um101 = um101[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "    tnf = pd.read_csv(\"path to tnf file\")\n",
    "    tnf = tnf[tnf[\"padj\"] <0.05]\n",
    "    tnf = tnf[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "    sf3030 = pd.read_csv(\"path to sf3030 file\")\n",
    "    sf3030 = sf3030[sf3030[\"padj\"] <0.05]\n",
    "    sf3030 = sf3030[[\"log2FoldChange\",\"padj\",\"Gene.name\"]]\n",
    "\n",
    "    pdgf = pd.read_csv(\"path to PDGF file\")\n",
    "    pdgf = pdgf[pdgf[\"padj\"] <0.05]\n",
    "\n",
    "    bvd = pd.read_csv(\"path to BVD file\")\n",
    "    bvd = bvd[bvd[\"padj\"] <0.05]\n",
    "\n",
    "    p450_median = p450.groupby('Gene.name').median().reset_index()\n",
    "    p900_median = p900.groupby('Gene.name').median().reset_index()\n",
    "    sf7009_median = sf7009.groupby('Gene.name').median().reset_index()\n",
    "    sf7044_median = sf7044.groupby('Gene.name').median().reset_index()\n",
    "    sf6222_median = sf6222.groupby('Gene.name').median().reset_index()\n",
    "    sf3030_median = sf3030.groupby('Gene.name').median().reset_index()\n",
    "    tnf_median = tnf.groupby('Gene.name').median().reset_index()\n",
    "    um101_median = um101.groupby('Gene.name').median().reset_index()\n",
    "    pdgf_median = pdgf.groupby('Gene.name').median().reset_index()\n",
    "    bvd_median = bvd.groupby('Gene.name').median().reset_index()\n",
    "\n",
    "    merged_table = pd.merge(p450_median, p900_median, on=\"Gene.name\", how=\"inner\")\n",
    "    p450_900_450 = merged_table[['Gene.name', 'padj_x', \"log2FoldChange_x\"]]\n",
    "    p450_900_900 = merged_table[['Gene.name', 'padj_y', \"log2FoldChange_y\"]]\n",
    "\n",
    "    merged_table['log2FoldChange'] = (merged_table['log2FoldChange_x'] + merged_table['log2FoldChange_y']) / 2\n",
    "    average_450_900 = merged_table[['Gene.name','log2FoldChange']]\n",
    "\n",
    "    average_merge_450_900 = corr_check.merge(average_450_900, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_450 = corr_check.merge(p450_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_900 = corr_check.merge(p900_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_sf7044 = corr_check.merge(sf7044_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_sf6222 = corr_check.merge(sf6222_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_sf7009 = corr_check.merge(sf7009_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_um101 = corr_check.merge(um101_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_sf3030 = corr_check.merge(sf3030_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_pdgf = corr_check.merge(pdgf_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_bvd = corr_check.merge(bvd_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "    merge_tnf = corr_check.merge(tnf_median, how='inner', on=None, left_on=symbol_col, right_on=\"Gene.name\", left_index=False, right_index=False)\n",
    "\n",
    "    print(merge_450)\n",
    "\n",
    "\n",
    "    log_fold_drug = \"log2FoldChange\"\n",
    "\n",
    "    try:\n",
    "        average_p900_p450_corr, _ = pearsonr(average_merge_450_900[disease_column], average_merge_450_900[log_fold_drug])\n",
    "    except:\n",
    "        average_p900_p450_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        p450_corr, _ = pearsonr(merge_450[disease_column], merge_450[log_fold_drug])\n",
    "    except:\n",
    "        p450_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        p900_corr, _ = pearsonr(merge_900[disease_column], merge_900[log_fold_drug])\n",
    "    except:\n",
    "        p900_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        sf7044_corr, _ = pearsonr(merge_sf7044[disease_column], merge_sf7044[log_fold_drug])\n",
    "    except:\n",
    "        sf7044_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        sf7009_corr, _ = pearsonr(merge_sf7009[disease_column], merge_sf7009[log_fold_drug])\n",
    "    except:\n",
    "        sf7009_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        sf6222_corr, _ = pearsonr(merge_sf6222[disease_column], merge_sf6222[log_fold_drug])\n",
    "    except:\n",
    "        sf6222_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        um101_corr, _ = pearsonr(merge_um101[disease_column], merge_um101[log_fold_drug])\n",
    "    except:\n",
    "        um101_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        tnf_corr, _ = pearsonr(merge_tnf[disease_column], merge_tnf[log_fold_drug])\n",
    "    except:\n",
    "        tnf_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        sf3030_corr, _ = pearsonr(merge_sf3030[disease_column], merge_sf3030[log_fold_drug])\n",
    "    except:\n",
    "        sf3030_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        pdgf_corr, _ = pearsonr(merge_pdgf[disease_column], merge_pdgf[log_fold_drug])\n",
    "    except:\n",
    "        pdgf_corr = \"NA\"\n",
    "\n",
    "    try:\n",
    "        bvd_corr, _ = pearsonr(merge_bvd[disease_column], merge_bvd[log_fold_drug])\n",
    "    except:\n",
    "        bvd_corr = \"NA\"\n",
    "\n",
    "    dict_of_correlations = {'Compound Dataset': [\"GEn-1124 Proteomics 450mg BID\", \"GEn-1124 Proteomics 900mg BID\",\n",
    "                                                \"GEn-1124 Proteomics Average\",\n",
    "                                               \"GEn-1124 RNA-Seq\",\n",
    "                                                 \"TNF RNA-Seq\",\n",
    "                                               \"GEn-3030 Proteomics\",\n",
    "                                               \"PDGF Proteomics\",\n",
    "                                               \"BVD Proteomics\"],\n",
    "                            'Total Compound Dataset Genes': [len(p450_median), len(p900_median),\n",
    "                            len(average_450_900), len(sf7044_median), len(tnf_median), len(sf3030_median), len(pdgf_median),len(bvd_median)\n",
    "                            ],\n",
    "                            'Overlapping Genes (Compound and Disease)': [len(merge_450), len(merge_900),\n",
    "                            len(average_merge_450_900), len(merge_sf7044), len(merge_tnf), len(merge_sf3030), len(merge_pdgf),len(merge_bvd)\n",
    "                            ],\n",
    "                            'Correlation with Disease': [p450_corr, p900_corr, average_p900_p450_corr, sf7044_corr,\n",
    "                                                         tnf_corr, sf3030_corr, pdgf_corr, bvd_corr\n",
    "                                                         ]}\n",
    "\n",
    "\n",
    "    correlation_table = pd.DataFrame.from_dict(dict_of_correlations)\n",
    "\n",
    "    return correlation_table\n",
    "\n",
    "\n",
    "\n",
    "def get_correlation_table(file):\n",
    "\n",
    "    symbol_col = \"Gene Name\"\n",
    "\n",
    "    log_col = 'logchange'\n",
    "    pval_col = 'pvalue'\n",
    "\n",
    "    corr_check = corr_check[corr_check[pval_col] <= 0.05]\n",
    "\n",
    "    corr_check = corr_check[((corr_check[log_col] <= -1) | (corr_check[log_col] >= 1))]\n",
    "\n",
    "    for i in corr_check.columns:\n",
    "        if i == symbol_col:\n",
    "            continue\n",
    "        else:\n",
    "            corr_check[i] = pd.to_numeric(corr_check[i], errors='coerce')\n",
    "\n",
    "    corr_check = corr_check.groupby(symbol_col).median().reset_index()\n",
    "\n",
    "    correlation_table = create_table(log_col, corr_check, symbol_col)\n",
    "    correlation_table\n",
    "\n",
    "    return correlation_table\n",
    "\n",
    "\n",
    "correlation_table = get_correlation_table(file_stream)\n",
    "\n",
    "print(correlation_table)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "id": "c9d0f2fd",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/cd/1pn_n69n5vzf7yz6_fmnrkjr0000gp/T/ipykernel_2798/2284541881.py:10: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  new_table.rename(columns={\"logchange\": 'ABMR_Log_Change', 'log2FoldChange': 'GEn1E-1124'}, inplace=True)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Expression Pattern  Kidney ABMR  GEn-1124 (RNA-Seq)\n",
      "Gene Name                                          \n",
      "ACKR1                  1.225526           -0.425452\n",
      "ANKRD1                 1.840000            0.749113\n",
      "APOBEC3G               1.020000           -1.213847\n",
      "CD37                   1.355402           -1.544997\n",
      "CD38                   1.316969           -0.478977\n",
      "CDH13                  1.056587            0.204116\n",
      "CORO1A                 1.364365           -0.686100\n",
      "CXCL10                 2.055996           -4.500982\n",
      "CXCL11                 2.269632           -2.565628\n",
      "GBP4                   1.273582           -2.734770\n",
      "GMFG                   1.033867            0.190516\n",
      "HCLS1                  1.006909           -0.424801\n",
      "IL7R                   1.256826           -0.390483\n",
      "IRF1                   1.013959            0.707819\n",
      "ITGA4                  1.309910            0.253935\n",
      "KLHL6                  1.022730            1.437805\n",
      "PDCD1LG2               1.015541           -0.145404\n",
      "SERPINB9               1.006601           -0.815805\n",
      "ST8SIA4                1.010000            1.294353\n",
      "STAT4                  1.040000            0.961017\n",
      "TAP1                   1.051252           -1.265112\n",
      "TNFRSF9                1.274744           -1.608895\n",
      "UBD                    1.290682           -3.350031\n",
      "VCAN                   1.030000            0.402082\n",
      "Expression Pattern  Kidney ABMR  GEn-1124 (RNA-Seq)\n",
      "Gene Name                                          \n",
      "CXCL11                 2.269632           -2.565628\n",
      "CXCL10                 2.055996           -4.500982\n",
      "ANKRD1                 1.840000            0.749113\n",
      "CORO1A                 1.364365           -0.686100\n",
      "CD37                   1.355402           -1.544997\n",
      "CD38                   1.316969           -0.478977\n",
      "ITGA4                  1.309910            0.253935\n",
      "UBD                    1.290682           -3.350031\n",
      "TNFRSF9                1.274744           -1.608895\n",
      "GBP4                   1.273582           -2.734770\n",
      "IL7R                   1.256826           -0.390483\n",
      "ACKR1                  1.225526           -0.425452\n",
      "CDH13                  1.056587            0.204116\n",
      "TAP1                   1.051252           -1.265112\n",
      "STAT4                  1.040000            0.961017\n",
      "GMFG                   1.033867            0.190516\n",
      "VCAN                   1.030000            0.402082\n",
      "KLHL6                  1.022730            1.437805\n",
      "APOBEC3G               1.020000           -1.213847\n",
      "PDCD1LG2               1.015541           -0.145404\n",
      "IRF1                   1.013959            0.707819\n",
      "ST8SIA4                1.010000            1.294353\n",
      "HCLS1                  1.006909           -0.424801\n",
      "SERPINB9               1.006601           -0.815805\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/cd/1pn_n69n5vzf7yz6_fmnrkjr0000gp/T/ipykernel_2798/2284541881.py:17: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  disease_columns[\"Expression Pattern\"] = \"Kidney ABMR\"\n",
      "/var/folders/cd/1pn_n69n5vzf7yz6_fmnrkjr0000gp/T/ipykernel_2798/2284541881.py:20: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  drug_columns[\"Expression Pattern\"] = name_of_drug\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def makeheatmap(table, drug_column, name_of_drug, top_number=None):\n",
    "    \n",
    "    table['mixlog'] = table['logchange'] + table[drug_column]\n",
    "    table = table.sort_values(\"logchange\", ascending=False)\n",
    "    \n",
    "#     Option to only keep the top 24 genes (top 12 upregulated and top 12 downregulated)\n",
    "\n",
    "#     table_head = table.head(12) \n",
    "#     table_tail = table.tail(12) \n",
    "#     table = pd.concat([table_head, table_tail])\n",
    "    \n",
    "    new_table = table[[\"Gene Name\", \"logchange\", \"log2FoldChange\"]]\n",
    "    new_table.rename(columns={\"logchange\": 'Disease_Log_Change', 'log2FoldChange': name_of_drug}, inplace=True)\n",
    "    \n",
    "    new_table.to_excel('path to create the raw file with the heatmap data for the compound', index=False)\n",
    "    \n",
    "    if top_number:\n",
    "        table = table.sample(top_number, random_state=42)       \n",
    "    disease_columns = table[[\"Gene Name\", \"logchange\"]]\n",
    "    disease_columns[\"Expression Pattern\"] = \"Disease Name\"\n",
    "\n",
    "    drug_columns = table[[\"Gene Name\", drug_column]]\n",
    "    drug_columns[\"Expression Pattern\"] = name_of_drug\n",
    "    drug_columns = drug_columns.rename(columns={drug_column:\"logchange\"})\n",
    "    drugname = name_of_drug\n",
    "\n",
    "    final_heatmap = pd.concat([disease_columns, drug_columns])    \n",
    "    \n",
    "    final_heatmap = final_heatmap.pivot(\"Gene Name\", \"Expression Pattern\",\"logchange\")\n",
    "    column_order = ['Disease Name', \"GEn-1124 (RNA-Seq)\"]\n",
    "    final_heatmap = final_heatmap.reindex(column_order, axis=1) \n",
    "    final_heatmap = final_heatmap.sort_values('Disease Name', ascending=False)\n",
    "\n",
    "    print(final_heatmap)\n",
    "\n",
    "    sns.heatmap(final_heatmap , cmap=\"coolwarm\", vmin=-2, vmax=+2)\n",
    "\n",
    "    \n",
    "# This function can be run on all drugs. Examples are below. Uncomment to use\n",
    "\n",
    "# average_p900_p450_heatmap = makeheatmap(average_merge_450_900, 'average', \"P_450_900_average\")\n",
    "# p900_p450_p450_heatmap = makeheatmap(merge_450_900_450, 'log2FoldChange_x', \"P_450_900_450\")\n",
    "# p900_p450_p900_heatmap = makeheatmap(merge_450_900_900, 'log2FoldChange_y', \"P_450_900_900\")\n",
    "# p450_heatmap = makeheatmap(merge_450, 'log2FoldChange', \"P450\")\n",
    "# p900_heatmap = makeheatmap(merge_900, 'log2FoldChange', \"P900\")\n",
    "# sf7044_heatmap = makeheatmap(merge_sf7044, 'log2FoldChange', \"GEn-1124 (RNA-Seq)\")\n",
    "# sf7009_heatmap = makeheatmap(merge_sf7009, 'log2FoldChange', \"SF7009\")\n",
    "# sf6222_heatmap = makeheatmap(merge_sf6222, 'log2FoldChange', \"SF6222\")\n",
    "# um101_heatmap = makeheatmap(merge_um101, 'log2FoldChange', \"UM101\")"
   ]
  }
 ],
 "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}