The identification of drug-target interactions (DTIs) is a key task in drug discovery. We propose an interpretable end-to-end deep learning framework to predict DTIs from low level representations. Experimental evaluation shows that this approach overall outperforms all baselines, and it is the only one capable of generalizing well to new drug targets. We also show the ability of our approach to provide biological insights to understand the nature of predicted DTIs.

Learning Objective 1: After participating in this session, the learner should be better able to:
- Recognize the benefits of in silico predicting drug-target interactions (DTIs)
- Compare current DTI prediction approaches
- Formulate an end-to-end approach to predict DTIs through deep neural representation
- Generate potential structure-level biological hypotheses of predicted DTIs via interpretable deep learning


Yingkai Gao, IBM
Achille Fokoue, IBM
Heng Luo, IBM T.J. Watson Research Center
Sanjoy Dey (Presenter)
IBM T.J. Watson Research Center

Arun Iyengar, IBM
Ping Zhang, IBM T.J. Watson Research Center

Presentation Materials: