Shared gene signatures and molecular mechanisms link ankylosing spondylitis and rheumatoid arthritis.

ABSTRACT: Ankylosing spondylitis (AS) and rheumatoid arthritis (RA) are closely related autoimmune diseases with shared mechanisms that remain unclear. This study aims to identify shared molecular signatures and hub genes underlying the co-occurrence of AS and RA using clinical and transcriptomic data, focusing on immune dysregulation pathways. The CBC data of 23,289 patients were collected, and six machine learning algorithms were applied to develop disease prediction models for AS and RA. Using permutation feature importance and Shapley Additive Explanations (SHAP) based on the optimal model, the top 10 features most influential for AS and RA prediction were identified, followed by selecting their intersections. Bioinformatics analysis was conducted to identify key immune cells associated with AS

SUBMITTER: Qin B 

PROVIDER: S-EPMC12217434 | biostudies-literature | 2025 Jul

REPOSITORIES: biostudies-literature