Dissecting the genetics of complex traits using summary association statistics

Summary association statistics from genome-wide association studies (GWAS) are widely available in large sample sizes across hundreds of complex traits. Analyses of such data can yield important insights, motivating the development of new statistical methods in this area.Single variant association analysis (including meta-analyses, conditional association and imputation) can be performed effectively using summary association data. These methods often rely on linkage disequilibrium (LD) information from population reference panels.Summary association data can be used to perform gene-based association tests to identify genes influencing complex traits. In particular, expression quantitative trait loci (eQTLs) can be integrated to identify genes whose expression levels influence complex traits, and rare variant association tests can aggregate evidence of association across multiple rare variants in a gene.Statistical fine-mapping of causal variant (or variants) at GWAS loci can be performed using summary association data, leveraging information on the strength of association, functional genomic annotations and differences in LD patterns across different populations.It is becoming increasingly clear that most complex traits and common diseases have a large number of causal variants with small effects. Summary association statistics can be used to understand these polygenic architectures and leverage them for polygenic risk prediction.Summary association statistics have broad utility in cross-trait analyses, including detecting pleiotropic effects and inferring genetic correlations between traits. Pleiotropic effects can be used in Mendelian randomization analyses to draw inferences about causal relationships among traits.