From a Multi-Omics Signature to a Therapeutic Candidate: Computational Prediction and Experimental Validation in Liver Fibrosis - PubMed
6 hours ago
- #Therapeutic Discovery
- #Machine Learning
- #Liver Fibrosis
- A robust six-gene signature (CLEC4M, COL25A1, ITGBL1, NALCN, PAPPA, PEG3) was developed using machine learning to predict advanced liver fibrosis across etiologies, achieving high diagnostic accuracy in internal and external validation.
- Single-cell RNA sequencing revealed that the signature genes are prominently enriched in fibroblast populations, highlighting their cellular context in liver fibrosis.
- Withaferin A (WFA) was identified through computational drug repurposing and validated as an anti-fibrotic agent; it attenuated fibrosis in vivo in a mouse model and suppressed hepatic stellate cell activation in vitro.