Proceeding talk – Theme: Systems.
Abstract
Clinical response to anti-cancer drugs varies between patients, partly due to differences in molecular alterations in the tumours. A classic approach to predict drug response is to combine these molecular alterations as features. However, this results in models mostly based on gene expression, which is difficult to interpret. To utilize all data types in a more balanced way, we developed TANDEM, a two-stage approach, which first explains response using upstream features (mutations, copy number, methylation and cancer type) and then explains the remainder using downstream features (gene expression). Applying TANDEM to 934 cell lines profiled across 265 drugs, we show that the resulting models are more interpretable, while retaining the same predictive performance as the classic approach. Interestingly, we find that response to MAPK pathway inhibitors is largely predicted by mutation data, while predicting response to DNA damaging agents requires gene expression data, in particular SLFN11 expression.
Authors
Nanne Aben, Netherlands Cancer Institute, Netherlands
Daniel J. Vis, Netherlands Cancer Institute, Netherlands
Magali Michaut, Netherlands Cancer Institute, Netherlands
Lodewyk F. A. Wessels, Netherlands Cancer Institute, Netherlands