Genetics

This blog post critically examines a recent mini-review on multiple sclerosis risk prediction, focusing on how polygenic risk scores and environmental risk scores are constructed, evaluated, and translated into practice. It explains why predictive performance is fundamentally limited by the polygenic architecture and bounded SNP-heritability of MS, and why commonly reported discrimination metrics can overstate real-world clinical usefulness in a low-prevalence disease. The discussion also addresses key barriers—including causal-variant uncertainty, rare-variant omission, gene–environment interactions, exposure misclassification, and cross-ancestry portability—that currently prevent individualized prediction from achieving clinical-grade accuracy. Finally, it outlines a realistic path forward, emphasizing prospective validation, improved modeling, and the most actionable near-term application: using risk stratification to enrich prevention trials and accelerate the evaluation of preventive interventions.