Benchmark on Drug Target Interaction Modeling from a Drug Structure Perspective

TL;DR

This paper provides a comprehensive benchmark of drug-target interaction models focusing on structural information, comparing GNN and Transformer methods, and proposing optimized model combinations that achieve state-of-the-art results.

Contribution

It offers a detailed comparison of explicit and implicit structure learning algorithms and introduces effective model combinations for improved performance.

Findings

GNN and Transformer-based models show competitive performance.

Optimized model combinations outperform individual models.

New state-of-the-art results achieved on multiple datasets.

Abstract

The prediction modeling of drug-target interactions is crucial to drug discovery and design, which has seen rapid advancements owing to deep learning technologies. Recently developed methods, such as those based on graph neural networks (GNNs) and Transformers, demonstrate exceptional performance across various datasets by effectively extracting structural information. However, the benchmarking of these novel methods often varies significantly in terms of hyperparameter settings and datasets, which limits algorithmic progress. In view of these, we conducted a comprehensive survey and benchmark for drug-target interaction modeling from a structural perspective via integrating tens of explicit (i.e., GNN-based) and implicit (i.e., Transformer-based) structure learning algorithms. We conducted a macroscopical comparison between these two classes of encoding strategies as well as the different featurization techniques that inform molecules' chemical and physical properties. We then carry out the microscopical comparison between all the integrated models across the six datasets via comprehensively benchmarking their effectiveness and efficiency. To ensure fairness, we investigate model performance under individually optimized configuration. Remarkably, the summarized insights from the benchmark studies lead to the design of model combos. We demonstrate that our combos can achieve new state-of-the-art performance on various datasets associated with cost-effective memory and computation.