PIGNet: A physics-informed deep learning model toward generalized drug-target interaction predictions
This addresses the challenge of generalization in in-silico drug discovery, offering improved accuracy for drug-target interaction predictions, though it is incremental as it builds on existing DNN approaches.
The authors tackled the problem of insufficient generalization in deep neural network-based drug-target interaction models by proposing PIGNet, which uses physics-informed equations and data augmentation to predict binding affinities, achieving outperforming docking and screening powers in the CASF 2016 benchmark.
Recently, deep neural network (DNN)-based drug-target interaction (DTI) models were highlighted for their high accuracy with affordable computational costs. Yet, the models' insufficient generalization remains a challenging problem in the practice of in-silico drug discovery. We propose two key strategies to enhance generalization in the DTI model. The first is to predict the atom-atom pairwise interactions via physics-informed equations parameterized with neural networks and provides the total binding affinity of a protein-ligand complex as their sum. We further improved the model generalization by augmenting a broader range of binding poses and ligands to training data. We validated our model, PIGNet, in the comparative assessment of scoring functions (CASF) 2016, demonstrating the outperforming docking and screening powers than previous methods. Our physics-informing strategy also enables the interpretation of predicted affinities by visualizing the contribution of ligand substructures, providing insights for further ligand optimization.