Drug–target interaction predictions with multi-view similarity network fusion strategy and deep interactive attention mechanism

• DOI: 10.1093/bioinformatics/btae346
• Publication Date: 2024-06-05T19:36:58Z
• Authors: Wei Song; Lewen Xu; Chenguang Han; Zhen Tian; Quan Zou
• Keywords: Discriminative model; Similarity (geometry); Fusion mechanism; Multilayer perceptron

Abstract Motivation Accurately identifying the drug–target interactions (DTIs) is one of crucial steps in drug discovery and repositioning process. Currently, many computational-based models have already been proposed for DTI prediction achieved some significant improvement. However, these approaches pay little attention to fuse multi-view similarity networks related drugs targets an appropriate way. Besides, how fully incorporate known interaction relationships accurately represent not well investigated. Therefore, there still a need improve accuracy models. Results In this study, we propose novel approach that employs Multi-view network fusion strategy deep Interactive mechanism predict Drug–Target Interactions (MIDTI). First, MIDTI constructs with their diverse information integrates effectively unsupervised manner. Then, obtains embeddings from multi-type simultaneously. After that, adopts interactive further learn discriminative comprehensively relationships. Finally, feed learned representations multilayer perceptron model underlying interactions. Extensive results indicate significantly outperforms other baseline methods on task. The ablation experiments also confirm effectiveness mechanism. Availability implementation https://github.com/XuLew/MIDTI.