MGACL: Prediction Drug–Protein Interaction Based on Meta-Graph Association-Aware Contrastive Learning

The identification of drug–target interaction (DTI) is crucial for drug discovery. However, how to reduce the graph neural network’s false positives due to its bias and negative transfer in the original bipartite graph remains to be clarified. Considering that the impact of heterogeneous auxiliary i...

Full description

Saved in:
Bibliographic Details
Published in:Biomolecules (Basel, Switzerland) Vol. 14; no. 10; p. 1267
Main Authors: Zhang, Pinglu, Lin, Peng, Li, Dehai, Wang, Wanchun, Qi, Xin, Li, Jing, Xiong, Jianshe
Format: Journal Article
Language:English
Published: Switzerland MDPI AG 01.10.2024
MDPI
Subjects:
Algorithms
Area Under Curve
Artificial intelligence
Associations
Comparative analysis
Connectivity
contrastive learning
Customization
Disease
Drug development
Drug discovery
Drug Discovery - methods
drug–target interaction
Genes
graph neural networks
Graphs
heterogeneous graph representation
Humans
Machine Learning
Neural networks
Neural Networks, Computer
Pharmaceutical Preparations - chemistry
Pharmaceutical Preparations - metabolism
Proteins - chemistry
Proteins - metabolism
R&D
Research & development
Semantics
Therapeutic targets
Transfer learning
China
contrastive learning
graph neural networks
drug–target interaction
heterogeneous graph representation
ISSN:2218-273X, 2218-273X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The identification of drug–target interaction (DTI) is crucial for drug discovery. However, how to reduce the graph neural network’s false positives due to its bias and negative transfer in the original bipartite graph remains to be clarified. Considering that the impact of heterogeneous auxiliary information on DTI varies depending on the drug and target, we established an adaptive enhanced personalized meta-knowledge transfer network named Meta Graph Association-Aware Contrastive Learning (MGACL), which can transfer personalized heterogeneous auxiliary information from different nodes and reduce data bias. Meanwhile, we propose a novel DTI association-aware contrastive learning strategy that aligns high-frequency drug representations with learned auxiliary graph representations to prevent negative transfer. Our study improves the DTI prediction performance by about 3%, evaluated by analyzing the area under the curve (AUC) and area under the precision–recall curve (AUPRC) compared with existing methods, which is more conducive to accurately identifying drug targets for the development of new drugs.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
These authors contributed equally to this work.
ISSN:2218-273X
2218-273X
DOI:10.3390/biom14101267