アブストラクト
| Date | 14:00-15:00 Apr 28, 2026 |
| Speaker | Aron Schulman Institute for Molecular Medicine Finland (FIMM), University of Helsinki, Finland |
| Title | An Attention-Based Method and Validation Guidelines for Drug-Target Interaction Prediction |
| Abstract |
Drug-target interaction (DTI) prediction is an active area of research aimed at reducing time and cost in drug discovery by enabling the characterization of proteome-wide binding profiles to support drug efficacy and safety. We have developed an attention-based deep learning method for predicting DTIs across seven target protein classes. We evaluate several molecular descriptors for drugs and targets and assess their generalizability across the target classes. Model performance is tested under different validation scenarios with independent test data to identify the most appropriate use cases. In related work, we examine trends in DTI prediction methods and provide guidelines for their validation and benchmarking.
|