onconpc_prediction_and_explanation_for_cup_tumors.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "4f19125c-7c88-41b6-8afd-4950f537f12c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'utils' from '/Users/intaemoon/Documents/github/onconpc/codes/utils.py'>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import joblib\n",
    "import sys\n",
    "sys.path.append('codes')\n",
    "import utils\n",
    "import importlib\n",
    "importlib.reload(utils)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4f3ab86b-c8be-45b1-a6b4-7c1c03b88e99",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[02:39:50] WARNING: /Users/travis/build/dmlc/xgboost/src/tree/./updater_quantile_hist.h:135: Attempted to load interal configuration for a model file that was generated by a previous version of XGBoost. A likely cause for this warning is that the model was saved with saveRDS() in R or pickle.dump() in Python. We strongly ADVISE AGAINST using saveRDS() or pickle.dump() so that the model remains accessible in current and upcoming XGBoost releases. Please use xgb.save() instead to preserve models for the long term. For more details and explanation, see https://xgboost.readthedocs.io/en/latest/tutorials/saving_model.html\n"
     ]
    }
   ],
   "source": [
    "# Load processed CUP data\n",
    "onconpc_processed_cups_df = pd.read_csv('data/onconpc_processed_cups_data.csv', index_col='RANDID')\n",
    "\n",
    "# Load fully trained OncoNPC model\n",
    "xgb_onconpc = joblib.load('model/xgboost_OncoNPC_full')\n",
    "\n",
    "# Specify cancer types to consider\n",
    "cancer_types_to_consider = ['Acute Myeloid Leukemia', 'Bladder Urothelial Carcinoma', 'Cholangiocarcinoma',\n",
    "                            'Colorectal Adenocarcinoma', 'Diffuse Glioma', 'Endometrial Carcinoma',\n",
    "                            'Esophagogastric Adenocarcinoma', 'Gastrointestinal Neuroendocrine Tumors', 'Gastrointestinal Stromal Tumor',\n",
    "                            'Head and Neck Squamous Cell Carcinoma', 'Invasive Breast Carcinoma', 'Melanoma', 'Meningothelial Tumor',\n",
    "                            'Non-Hodgkin Lymphoma', 'Non-Small Cell Lung Cancer', 'Ovarian Epithelial Tumor', 'Pancreatic Adenocarcinoma',\n",
    "                            'Pancreatic Neuroendocrine Tumor', 'Pleural Mesothelioma', 'Prostate Adenocarcinoma', 'Renal Cell Carcinoma',\n",
    "                            'Well-Differentiated Thyroid Cancer']\n",
    "\n",
    "# Predict primary sites of CUP tumors\n",
    "cup_preds_df = utils.get_xgboost_cancer_type_preds(xgb_onconpc,\n",
    "                                                   onconpc_processed_cups_df,\n",
    "                                                   cancer_types_to_consider)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d7d21a70-6874-4cc7-ad75-6a8cb5dd9617",
   "metadata": {},
   "outputs": [],
   "source": [
    "shaps_cup = utils.obtain_shap_values(xgb_onconpc, onconpc_processed_cups_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cbf44d50-a887-46b1-bad6-81226fd18559",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Specify query tumor sample ID\n",
    "query_randid = 'DFCI_963265'\n",
    "\n",
    "# Get OncoNPC prediction\n",
    "pred_prob = cup_preds_df.at[query_randid, 'max_posterior']\n",
    "pred_cancer = cup_preds_df.at[query_randid, 'cancer_type']\n",
    "pred_cancer_idx = cancer_types_to_consider.index(pred_cancer)\n",
    "\n",
    "# Get SHAP-based explanation for the prediction\n",
    "feature_sample_df = onconpc_processed_cups_df.loc[query_randid]\n",
    "shap_pred_cancer_df = pd.DataFrame(shaps_cup[pred_cancer_idx],\n",
    "                                   index=onconpc_processed_cups_df.index,\n",
    "                                   columns=onconpc_processed_cups_df.columns)\n",
    "shap_pred_sample_df = shap_pred_cancer_df.loc[query_randid]\n",
    "\n",
    "# Generate explanation plot\n",
    "sample_info = f'RANDID: {query_randid}\\nPrediction: {pred_cancer}\\nPrediction probability: {pred_prob:.3f}'\n",
    "feature_group_to_features_dict = utils.partiton_feature_names_by_group(onconpc_processed_cups_df.columns)\n",
    "utils.get_individual_pred_interpretation(shap_pred_sample_df, feature_sample_df, feature_group_to_features_dict,\n",
    "                                         sample_info=sample_info)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6cec33c-b472-4d9e-ac9c-a57badb7da9c",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}