Integrative Machine Learning and Structural Modeling Identify Multitarget Therapeutic Candidates for Idiopathic
Ran Ding1,2,3,4, Yuan Zhang1,2,3,4
1School of Biomedical Sciences and Engineering, South China University of Technology, Guangzhou International Campus, Guangzhou 511442, P. R. China.
ACS Omega
|April 13, 2026
Summary
Researchers developed a computational framework to identify new therapies for idiopathic pulmonary fibrosis (IPF). This approach prioritizes drug candidates targeting multiple pathways involved in this progressive lung disease.
Area of Science:
- Pulmonary Medicine
- Computational Biology
- Pharmacology
Background:
- Idiopathic pulmonary fibrosis (IPF) is a progressive, fatal lung disease with complex cellular heterogeneity.
- Dysregulated signaling networks in IPF present significant challenges for developing effective therapeutics.
Purpose of the Study:
- To develop an integrative computational framework for prioritizing multitarget therapeutic candidates for IPF.
- To combine machine learning, transcriptomics, genetic inference, and molecular modeling for drug discovery in IPF.
Main Methods:
- Machine learning models predicted compounds targeting 13 IPF-associated receptor tyrosine kinases.
- Single-cell RNA sequencing identified target enrichment in relevant cell populations.
- Mendelian randomization and molecular simulations assessed genetic links and ligand-target interactions.
Main Results:
- Clinically used kinase inhibitors were identified as potential multitarget candidates.
- PDGFRB was highlighted as a gene causally associated with IPF susceptibility.
- Stable interactions were predicted between candidate ligands and PDGFRB.
Conclusions:
- The integrative framework systematically prioritizes multitarget therapeutics for IPF.
- This approach links computational predictions with biological and genetic evidence.
- The strategy aims to identify candidates modulating multiple disease pathways in IPF.
