|
Abstract:
|
I discuss multivariate approaches for aggregate brain imaging data with genetic data sets including SNPs and epigenetics (methylation). First, in a parallel ICA (pICA) framework, we derived imaging-genetic patterns pairing structural/functional and genetic and epigenetic combinations.Next we calculated an aggregated risk score weighting the contributing genotypes by the disease risk. PICA indicated several loci whose aggregated effects contributed to gray matter differences in a network including the superior temporal gyrus. Methylation showed much stronger links to structural MRI. The aggregated psychiatric genomic consortium risk score also shows relationships to large scale gray matter networks and to dynamic functional network connectivity. Our results are consistent with the idea that just as schizophrenia is a polygenic disorder, the imaging deficits that contribute to it are also polygenic. The ability to examine the combined effects of multiple small genetic effects allows us to assess genetic pathways for potential mechanisms of multiple patterns of brain imaging effects in schizophrenia.
|
