CDRP chemical and phenotypic space notebook - SF4

broadinstitute · Jan 8, 2024 · ce6a220 · ce6a220
1 parent 17662e9
commit ce6a220
Showing 1 changed file with 199 additions and 0 deletions.
diff --git a/cdrp/Chemical-and-phenotypic-spaces.ipynb b/cdrp/Chemical-and-phenotypic-spaces.ipynb
@@ -0,0 +1,199 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import pandas as pd\n",
+    "import seaborn as sb\n",
+    "import umap"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "main = pd.read_csv('data/treatment_level_aux_combined.csv.gz')\n",
+    "columns = [str(i) for i in range(672)]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fingerprints = np.load('data/fingerprints_cdrp.npz')['features']\n",
+    "cdrp_smiles_scaffolds = pd.read_csv('data/cdrp_smiles_scaffolds.csv')\n",
+    "Y = pd.read_csv(\"data/CDRP_MOA_MATCHES_official.csv\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#get new UMAP embeddings\n",
+    "reducer = umap.UMAP()\n",
+    "embeddings = reducer.fit_transform(fingerprints)\n",
+    "print(fingerprints.shape, embeddings.shape)\n",
+    "aux = pd.concat((pd.DataFrame(embeddings, columns=[\"UMAP 1\", \"UMAP 2\"]), cdrp_smiles_scaffolds.reset_index()), axis=1)\n",
+    "#aux\n",
+    "aux = pd.merge(aux, Y, left_on = 'Metadata_BROAD_ID', right_on = 'Var1', how = 'left')\n",
+    "\n",
+    "#to read aux used in publication uncomment next line\n",
+    "#aux = pd.read_csv('data/chemical_aux_umap.csv')\n",
+    "\n",
+    "#UMAP embeddings that were used for the supplementary figure are already in this repository\n",
+    "#aux.to_csv('data/chemical_aux_umap.csv', index = False)\n",
+    "\n",
+    "sb.scatterplot(data=aux, x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"lightpink\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "moas = []\n",
+    "for k,r in Y.iterrows():\n",
+    "    for i in r[\"Metadata_moa.x\"].split(\"|\"):\n",
+    "        moas.append(i)\n",
+    "\n",
+    "moas = pd.DataFrame({'MoA': moas })"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig = plt.figure(figsize=(10,10))\n",
+    "a  = \"lipoxygenase inhibitor\"\n",
+    "g = sb.scatterplot(data=aux[~aux['Metadata_moa.x'].str.contains(a)], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"dodgerblue\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "h = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(a)], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"limegreen\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
+    "x_lims = (None, None)\n",
+    "y_lims = (None, None)\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "selected_moas = ['adenosine receptor agonist', 'adrenergic receptor antagonist', 'dopamine receptor agonist', 'egfr inhibitor', \n",
+    "                 'estrogen receptor agonist', 'glucocorticoid receptor agonist', \"tyrosine kinase inhibitor\",\n",
+    "                 'opioid receptor antagonist', \"bacterial dna gyrase inhibitor\", \"hmgcr inhibitor\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig = plt.figure(figsize=(10,10))\n",
+    "h = sb.scatterplot(data=aux[~aux['Metadata_moa.x'].str.contains('|'.join(selected_moas))], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"dodgerblue\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "u = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[0])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"mediumorchid\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "v = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[1])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"indigo\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "w = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[2])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"teal\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "x = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[3])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"limegreen\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "y = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[4])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"gold\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "z = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[5])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"blue\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "k = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[6])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"salmon\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "l = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[7])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"rosybrown\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "m = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[8])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"hotpink\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "n = sb.scatterplot(data=aux[aux['Metadata_moa.x'].str.contains(selected_moas[9])], x=\"UMAP 1\", y=\"UMAP 2\", s=100, color=\"crimson\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "\n",
+    "plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
+    "x_lims = (None, None)\n",
+    "y_lims = (None, None)\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig.savefig(\"chemical_space_moa.png\") \n",
+    "fig.savefig(\"chemical_space_moa.svg\") "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig = plt.figure(figsize=(10,10))\n",
+    "h = sb.scatterplot(data=main[~main['Metadata_moa.x'].str.contains('|'.join(selected_moas))], x=\"X\", y=\"Y\", s=100, color=\"dodgerblue\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "u = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[0])], x=\"X\", y=\"Y\", s=100, color=\"mediumorchid\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "v = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[1])], x=\"X\", y=\"Y\", s=100, color=\"indigo\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "w = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[2])], x=\"X\", y=\"Y\", s=100, color=\"teal\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "x = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[3])], x=\"X\", y=\"Y\", s=100, color=\"limegreen\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "y = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[4])], x=\"X\", y=\"Y\", s=100, color=\"gold\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "z = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[5])], x=\"X\", y=\"Y\", s=100, color=\"blue\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "k = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[6])], x=\"X\", y=\"Y\", s=100, color=\"salmon\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "l = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[7])], x=\"X\", y=\"Y\", s=100, color=\"rosybrown\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "m = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[8])], x=\"X\", y=\"Y\", s=100, color=\"hotpink\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "n = sb.scatterplot(data=main[main['Metadata_moa.x'].str.contains(selected_moas[9])], x=\"X\", y=\"Y\", s=100, color=\"crimson\", linewidth=0.5, edgecolor=\"black\", alpha=0.8)\n",
+    "\n",
+    "plt.legend(bbox_to_anchor=(1.05, 1), loc=2, borderaxespad=0.)\n",
+    "x_lims = (None, None)\n",
+    "y_lims = (None, None)\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig.savefig(\"phenotypic_space_moa.png\") \n",
+    "fig.savefig(\"phenotypic_space_moa.svg\") "
+   ]
+  }
+ ],
+ "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": 4
+}