Tumor Drug Sensitivity Prediction Project

Overview

This project replicates and extends the study presented in "Predicting Tumor Response to Drugs Based on Gene-Expression Biomarkers of Sensitivity Learned from Cancer Cell Lines". The study aims to predict tumor response to drugs using gene-expression biomarkers learned from cancer cell lines, primarily by leveraging a Genetic Algorithm (GA) for feature selection and k-Nearest Neighbors (KNN) for prediction. This project allows users to predict drug sensitivity based on gene-expression data.

Requirements

• Python 3.8 or above
• Recommended hardware: CPU or GPU if large datasets are used
• Recommended OS: Ubuntu, MacOS, or Windows

Set up environment

• Clone this repository and navigate to the project directory.
• Install dependencies using:

  pip install -r requirements.txt

Download Datasets

• TCGA RNA-seq: https://gdc.cancer.gov/about-data/publications/pancanatlas (filename: EBPlusPlusAdjustPANCAN_IlluminaHiSeq_RNASeqV2.geneExp.tsv)
  • Scroll to "Supplemental Data"
  • Download RNA (Final)
• GTEx RNA-seq: https://www.gtexportal.org/home/datasets (filename: GTEx_Analysis_2017-06-05_v8_RNASeQCv1.1.9_gene_tpm.gct.gz)
  • Select data type: "Bulk Tissue Expression"
  • Navigate to GTEx Analysis V8
  • File description: Gene TPMs
• CCLE RNA-seq: https://depmap.org/portal/data_page/?tab=allData (filename: CCLE_RNAseq_rsem_genes_tpm_20180929.txt.gz)
  • Push DepMap Public, scroll to "Other Omics," and find "CCLE 2019"
  • File description: CCLE RNAseq gene expression data fpr 1019 cell lines (RPKM)
• Drug Sensitivity (IC50 values): https://www.cancerrxgene.org/downloads/bulk_download (filenames: GDSC1_fitted_dose_response_17Jul19.txt & GDSC2_fitted_dose_response_15Oct19.txt)
  • Scroll to “Archive,” rather than clicking GDSC1 and GDSC2 under "Screening Data"
  • GDSC1 and GDSC2 are updated frequently, so the versions listed above are now in the archive
  • To navigate and download data from the FTP site use the FileZilla Client - https://filezilla-project.org/
    • Host: ftp.sanger.ac.uk
    • Leave all else blank, and push quick connect
  • Navigate through folders to enter pub/project/cancerrxgene/releases/
    • Enter release-8.0 and download GDSC1_fitted_dose_response_17Jul19.csv
    • Enter release-8.1 and download GDSC2_fitted_dose_response_15Oct19.csv