Comprehensive Analysis of Bioinformatics and Experimental Validation Reveals New Diagnostic Markers and Potential Drug Targets for Osteoarthritis

Multi-model machine learning identifies MGAT4C, COL8A1, and NCAM1 as novel diagnostic biomarkers and therapeutic targets for OA.An interpretable artificial neural network diagnostic model is constructed, significantly improving OA diagnosis accuracy.Experimental validation confirms MGAT4C is significantly upregulated in OA cartilage, and its knockdown alleviates chondrocyte catabolism.CMap-based reverse screening predicts and verifies small-molecule compounds capable of reversing OA phenotypes.This study provides integrated multi-omics evidence for precise OA diagnosis and targeted therapy.