Network-based meta-analysis maps stage-dependent molecular programs in MASLD through MASLD-META NETWORK application

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the leading cause of chronic liver pathologies worldwide, with a growing clinical burden. Diagnosing MASLD currently relies on liver biopsy, which restricts early detection and the ability to observe molecular changes throughout disease progression. To address this, recent research has created databases for searching genes and predicting multi-gene signatures for disease progression.

However, a requirement remains for an interactive and comprehensive meta-analysis of MASLD patient datasets that include histological metadata.

In this study, researchers conducted a meta-analysis of RNA-seq datasets using the NAFLD Activity Score (NAS) in 897 patients and fibrosis stage in 856 patients. Comparisons were made across histological stages to identify differentially expressed genes associated with disease progression. The findings were subsequently made available through the MASLD-META NETWORK, a dedicated web server (https://masld.scilicium.com) that allows users to explore the meta-analysis results interactively across various network modalities.

To further characterize gene expression dynamics across increasing disease stages, the researchers employed Louvain clustering. Network-based parameters, including centrality in combination with meta-analysis scores, were utilized to highlight central genes and pathways implicated in disease mechanisms. As a result, the MASLD-META NETWORK identified COL1A1, COL3A1, THBS2, FBLN5, and PDGFA as the most central genes, while SULF2, MMP14, IL32, GPNMB, and COL3A1 were identified as potential markers of earlier transcriptional alterations.

Additionally, network analysis revealed LAMA2 and LAMA3 as previously unrecognized central candidate targets.