Daniël J. Vis

The paper shows that drug response predictors trained on perturbation response model-generated profiles outperform those trained on baseline or pre-treatment profiles.

The paper develops TRANSACT, a transfer learning framework for predicting clinical drug response, outperforming existing methods on multiple datasets.

The paper uses multi-task learning to identify biomarkers of anti-cancer drug synergy using low-sample-size drug combination data from 85 cell lines and 167 drug combinations.