Genetics

This post covers Noroozi and colleagues' 2026 medRxiv preprint using layered Mendelian randomization to link lipid metabolism to multiple sclerosis severity. It notes that only one metabolite in the primary screen clears the study's own pre-specified threshold, shows that the pathway-based multivariable analysis is where arachidonate and betaine reach real significance, explains why colocalization at the chromosome 11 FADS locus cannot separate FADS1 from three neighbouring genes, sets out the functional variant and expression data that carry the gene assignment instead, covers the brain cell types where reduced FADS1 tracks with severity, and closes on the CYP4F2 signal and the study that would test the whole argument in patients.