TCoCPIn reveals topological characteristics of chemical protein interaction networks for novel feature discovery

Abstract Understanding chemical–protein interactions (CPIs) is crucial for drug discovery and biological research, yet their complexity often challenges traditional methods. We propose TCoCPIn, a novel framework integrating graph neural networks (GNN) with the comprehensive topological characteristi...

Full description

Saved in:
Bibliographic Details
Main Authors: Jianshi Wang, Yukio Ohsawa
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-01410-7
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Abstract Understanding chemical–protein interactions (CPIs) is crucial for drug discovery and biological research, yet their complexity often challenges traditional methods. We propose TCoCPIn, a novel framework integrating graph neural networks (GNN) with the comprehensive topological characteristics index (CTC), which combines multiple topological metrics to enhance predictive accuracy. TCoCPIn significantly outperforms traditional and embedding-based methods, achieving higher accuracy, precision, and recall in CPI prediction. A case study highlights its ability to predict potential interactions, such as between ibuprofen and TNF-alpha, demonstrating its utility in identifying novel therapeutic targets. These findings illustrate TCoCPIn’s potential to uncover hidden associations and key nodes in CPI networks, providing new opportunities for drug discovery and disease mechanism exploration. Future work will focus on experimental validation and expanding the application of TCoCPIn to other biological systems.
ISSN:2045-2322