Genetics

Three recent multiple sclerosis papers of the same kind, public datasets run through a statistical pipeline with single-cell or single-nucleus sequencing in the middle, read for the point where the sign of an effect reverses between layers. Huang and colleagues built a causal chain in which serotonin raises MS risk, ADA lowers it and serotonin lowers ADA, then found ADA highest in T-NK cells but higher in the low immune receptor signalling group, and noted themselves that progression was assessed through cross-sectional case-control data. Qiao and colleagues reached GPX1 and RCN1 in MS brain tissue, both raised in human tissue and in mice, after which single-nucleus data showed the two genes moving in opposite directions across pericytes, microglia, fibroblasts, oligodendrocyte precursors and endothelial cells, and the Mendelian randomisation step returned a protective odds ratio for one of them. Wang and colleagues found ROMO1 the only gene selected across rheumatoid arthritis, MS and type 1 diabetes, with monocyte correlations of 0.20, 0.41 and minus 0.43 and odds ratios of 0.52, 3.42 and a non-significant 0.87. The close sets out which layer each sign belongs to.