scRNA-seq Drug Repurposing Pipeline
Semi-supervised ML pipeline for drug candidate identification in cancer immunotherapy using single-cell RNA-seq data
Overview
At Huntsman Cancer Institute, I developed a comprehensive computational pipeline for identifying drug candidates targeting resistant tumor subpopulations in cancer immunotherapy.
Key Features
- scDrug Integration: Applied scDrug to pre-treatment ESCC (Esophageal Squamous Cell Carcinoma) scRNA-seq data to identify drug candidates
- GDSC-based Predictions: Predicted cell-death percentages using Genomics of Drug Sensitivity in Cancer (GDSC) models for ICI combination strategies
- Semi-supervised ML Pipeline: Built a pipeline to track treatment resistance using scRNA-seq data
- Therapy Validation: Predicted and validated candidate therapies using iDEA, integrating DSigDB and DREIMT databases
Technologies Used
- Languages: Python, R
- ML Frameworks: Scikit-learn, PyTorch
- Bioinformatics: Seurat, scDrug, iDEA
- Databases: DSigDB, DREIMT, GDSC
Impact
This work contributed to a publication in Pharmaceuticals (2025) on harnessing single-cell RNA-seq for computational drug repurposing in cancer immunotherapy.