Integrating AI-enhanced kinase enrichment analysis (KEA) with geometric deep learning and federated learning for precision drug repurposing

2026-04-30

PRE

Saleem Iqbal *

Jing Chen

Debnath Pal

Bairong Shen *

DOI：10.1016/j.drudis.2026.104687

Original

Original request for help

Save

Abstract

En 中文

• Introduces an integrative framework combining Kinase Enrichment Analysis (KEA) with geometric deep learning and federated learning for precision drug repurposing. • Demonstrates how AI-guided KEA bridges systems-level phosphoproteomics and atomic-resolution modeling to uncover “dark kinases” across diseases such as Alzheimer’s and cancer. • Validates the KEA–Kinhibit pipeline through a proof-of-concept Alzheimer’s disease case study, achieving 80% top-k precision and 4.2-fold enrichment in kinase inhibitor prediction. • Highlights the role of explainable and generative AI in PROTAC design, quantum machine learning, and 3D structure-based drug discovery. • Establishes a privacy-preserving, federated AI framework for secure multi-institutional learning, promoting ethical and collaborative drug discovery. • Proposes a dynamic, AI-enhanced KEA platform as a next-generation tool for translational pharmacology, bridging omics data to actionable therapeutics.

Keywords:

Kinase Enrichment Analysis

Geometric Deep Learning

Federated Learning

Drug Repurposing

Precision Medicine

Integrating AI-enhanced kinase enrichment analysis (KEA) with geometric deep learning and federated learning for precision drug repurposing

Abstract

Journal

Organization

Cited Papers

Citing Papers