One way to approach this challenge is through gene-based aggregation methods that can increase power to detect associations beyond genome-wide significant loci and yield more biologically relevant information. In other words, rather than studying individual risk variants, these data can be collapsed at the level of genes to reduce multiple-testing burden and resolve key disorder-associated biological processes. For example, transcriptome wide association studies (TWAS) achieve this by integrating mRNA expression data with GWAS association data to detect genes for which genetically predicted expression is associated with the trait (Wainberg et al., 2019). TWAS can be conceptualised as a genetic approach to more traditional differential expression analyses. Specifically, rather than directly measuring mRNA expression in cases and controls, estimates of genetic effects on mRNA expression are integrated with the effect of those same genetic variants on a phenotype or disorder of interest. The expression component of this approach constructs a model that predicts mRNA expression for each gene using genetic variants. As a result, TWAS is capable of prioritising genes that may be involved in the disorder and assign them a
