Pranjal Gehlot
BME MS Thesis Defense Presentation
Date: 2026-04-17
Time: 11:30 AM
Location / Meeting Link: EBB 5029

Committee Members:
Dr. S Balakrishna Pai; Dr. May Dongmei Wang; Dr. Ahmet Coskun


Title: Structure-guided Genetics-informed Embeddings Enable Zero-shot Drug Sensitivity Prediction

Abstract:
Accurate prediction of anticancer drug response remains a central challenge in precision oncology, as experimental screening across diverse cell lines is costly and time-intensive. Large-scale pharmacogenomic resources such as the Genomics of Drug Sensitivity in Cancer enable systematic analysis of drug–cell line interactions; however, most computational approaches require response data for each drug, limiting generalization to novel compounds. Here, we propose a graph neural network framework that learns drug representations directly from molecular structure, supervised by pharmacogenomic similarity derived from response profiles. By aligning structural embeddings with biological response similarity, the model captures meaningful relationships between drugs while remaining independent of cell line features at inference. This enables zero-shot (cold-start) prediction, allowing similarity estimation and downstream prioritization for previously unseen drugs using only their molecular graphs. Our model was evaluated on the data from the Genomics of Drug Sensitivity in Cancer (GDSC) datasets accessed via PyTDC v1.1.15. It achieves substantial improvements in functional similarity prediction (Pearson=0.423, NDCG@10=0.480) and enables structure-only zero-shot twin matching of sensitive cell lines (average 53% hit rate, up to 100% in top cases), demonstrating its utility for guiding preclinical experimental design in precision oncology.