- Three-gene signature identified: EPYC, MAGED1, and LAP3 were consistently selected as overlapping features from two machine learning approaches, strengthening confidence in their selection.
- Good diagnostic performance in datasets: Each gene showed strong ability to distinguish RA from OA (AUC > 0.85) in two independent transcriptomic datasets.
- Immune-related biology: Differentially expressed genes in RA were enriched in immune pathways (e.g., T-cell activation), and EPYC and LAP3 were associated with multiple immune cell populations.
- Experimental support (limited): In a TNF-α–stimulated RA cell model, EPYC and LAP3 expression increased while MAGED1 decreased, consistent with computational findings.
- Not clinically ready: Small sample sizes, cross-sectional design, and validation limited to mRNA in a single cell model mean the findings remain preliminary and require external clinical validation.
Daily News
Stay up to date with the latest clinical headlines and other information tailored to your specialty.
