hyperiDPath: A multilayer hypergraph drug repurposing path-reasoning framework based on hypergraph learning

Libing Bai

doi:10.1016/j.neucom.2026.135201

hyperiDPath: A multilayer hypergraph drug repurposing path-reasoning framework based on hypergraph learning

2026-09-22

0

PRE

AI

DOI：10.1016/j.neucom.2026.135201

Original

Original request for help

Save

Abstract

En 中文

• We introduce a novel multilayer hypergraph model to incorporate higher-order interactions in biomolecular networks, integrating transcription factors, genes, proteins, and chemicals. • We design a specialized attention module within a deep learning framework, hyperiDPath, to extract key drug-disease mechanism insights based on these multilayer hypergraph interactions. • Extensive comparative experiments on real-world datasets demonstrate the superior performance of hyperiDPath over state-of-the-art benchmarks in drug repurposing, while also aligning predicted pathways with clinical evidence.

Keywords:

Biomolecular multilayer hypergraph

Drug repurposing

Hypergraph deep learning

Mechanism of drug action

hyperiDPath: A multilayer hypergraph drug repurposing path-reasoning framework based on hypergraph learning

Abstract

Journal

Organization

Cited Papers