Three Genes May Help Distinguish RA From OA
Machine learning and cell-line validation identify candidate biomarkers for differentiating rheumatoid arthritis vs osteoarthritis
By
Andrea Surnit
April 22, 2026
Objective: To identify potential genetic biomarkers that can differentiate rheumatoid arthritis (RA) from osteoarthritis (OA).
Approach: Key Findings: EPYC, MAGED1, and LAP3 are potential biomarkers for distinguishing RA from OA. Diagnostic performance of the identified genes was strong, with AUC values above 0.85. The support vector machine model achieved approximately 93% accuracy. Interpretation: The findings suggest that the identified genes may reflect immune activity in joint tissue, aiding in the differentiation of RA from OA.
Limitations: Small sample sizes and reliance on cross-sectional transcriptomic data limit causation establishment. Experimental validation was restricted to messenger RNA expression in a single cell model. Drug sensitivity analyses were based on cancer cell data, which may not accurately represent RA responses. Conclusion: Further validation in independent clinical cohorts and mechanistic studies is necessary before clinical application of these biomarkers.
Sources: