Genetics

This post covers Gourraud and colleagues' 2013 Brain study, which turned the three-dimensional distribution of white matter lesions into a quantitative phenotype for GWAS. It follows the registration groundwork that makes voxel-wise comparison possible, the lesion probability map differences by HLA-DRB1*15:01 carriage and by sex, why voxel-wise association fails and what the cluxel trait and principal component decomposition replace it with, the 31 associations carried by the eighth component with SYK and TRAPPC9 among them, the 48-gene interaction module enriched for myelination and axon guidance, and the authors' own account of what independent confirmation would require.