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

Abstract

Motivation: Accurately identifying the drug-target interactions (DTIs) is one of the crucial steps in the drug discovery and drug repositioning process. Currently, many computational-based models have already been proposed for DTI prediction and achieved some significant improvement. However, these approaches pay little attention to fuse the multi-view similarity networks related to drugs and targets in an appropriate way. Besides, how to fully incorporate the known interaction relationships to accurately represent drugs and targets is not well investigated. Therefore, there is still a need to improve the accuracy of DTI prediction models.

Results: In this study, we propose a novel approach that employs Multi-view similarity network fusion strategy and deep Interactive attention mechanism to predict Drug-Target Interactions (MIDTI). First, MIDTI constructs multi-view similarity networks of drugs and targets with their diverse information and integrates these similarity networks effectively in an unsupervised manner. Then, MIDTI obtains the embeddings of drugs and targets from multi-type networks simultaneously. After that, MIDTI adopts the deep interactive attention mechanism to further learn their discriminative embeddings comprehensively with the known DTI relationships. Finally, we feed the learned representations of drugs and targets to the multilayer perceptron model and predict the underlying interactions. Extensive results indicate that MIDTI significantly outperforms other baseline methods on the DTI prediction task. The results of the ablation experiments also confirm the effectiveness of the attention mechanism in the multi-view similarity network fusion strategy and the deep interactive attention mechanism.

Availability and implementation: https://github.com/XuLew/MIDTI.

Figure 1.

The overall framework of MIDTI. In Step 1, MIDTI constructs the integrated similarity networks of drugs and targets with their multisource information, as well as the drug–target bipartite network and the drug–target heterogeneous network. In Step 2, MIDTI learns the embeddings of drugs and targets from multiple networks respectively. In Step 3, MIDTI adopts the deep interactive attention mechanism to learn discriminative representations of drugs and targets. In Step 4, MIDTI predicts the potential DTIs with the MLP classifier.

Figure 2.

The four steps of multi-view drug similarity network fusion strategy. Step 1: Take different similarity networks of drugs as input and learn the embeddings of drugs from different networks. Step 2: Integrate the embeddings of drugs with the multi-view attention mechanism. Step 3: Reconstruct the integrated drug network through dot product operation on the integrated drug features. Step 4: Train MIDTI by minimizing reconstruction error between the reconstructed network and each original drug similarity network.

Figure 3.

Three types of mechanism in deep interactive attention module. (A) SA mechanism, (B) DTA mechanism, (C) TDA mechanism.

Figure 4.

The deep interactive attention mechanism based on a cascade of interactive attention layers. Each interactive attention layer contains the corresponding SA, DTA, and TDA mechanisms respectively.

Figure 5.

The results of MIDTI as well as other baseline approaches on AUC and AUPR metrics.

Figure 6.

Visualization of the learned drug–target embeddings by MIDTI under different epochs.