Dissect the genetic architecture for sociological traits through integrative analysis of GWAS and functional annotations

Genome-wide association studies (GWAS) have identified tens of thousands of associations for numerous complex traits. However, despite the identifications of associated genetic variants, interpretation of GWAS findings remains challenging. The complex structure of linkage disequilibrium in the human genome, coupled with weak effect sizes of common genetic variants, hinder our ability to identify biologically functional genetic variants and understand their functional mechanism. Recent advances in epigenetic and transcriptomic functional annotations have accelerated discoveries in a variety of human genetics applications including GWAS downstream analysis. In this project, we leverage integrative genomic functional annotations in GWAS data to dissect the genetic architecture of complex traits. Specifically, we will integrate the requested GWAS data with epigenetic and transcriptomic annotation data in public repositories (e.g. Epigenomics Roadmap Project, ECNODE, and GTEx) to explore the underlying genetic architecture of various sociogenomics traits available in the HRS, examine shared genetic components among these traits, leverage pleiotropy and functional annotation information to prioritize genetic variants affecting these phenotypes, and robust and interpretable produce genetic prediction models. We think that integrating functional annotation information can effectively reduce noises and spurious associations in the non-functional regions in the human genome. More importantly, the tissue-specific nature of epigenetic and transcriptomic data would provide novel insights into the genetic basis and functional pathways of sociogenomic phenotypes. Finally, using better prioritized variants and annotation-informed effect size estimates
can improve the prediction accuracy of polygenic risk score, which enhances the statistical power in studying the genetic relationship among multiple phenotypes.

Overwhelming evidence indicates that common genetic variants account for a substantial proportion of phenotypic variance in many complex behavioral phenotypes. As a systematic and robust approach, GWAS can effectively
identify genetic variants associated with human traits. In this study, we employ genetic data from HRS to identify genetic variants associated with a variety of sociological phenotypes. Then, we will apply state-of-the-art
statistical and computational methods to help interpret our findings. Specifically, we will integrate external annotation information of the human genome to fine-map causal variants at identified genetic loci, identify
related tissue and cell types for sociological traits, and identify candidate risk genes. Further, by jointly modeling multiple traits, we dissect the shared and distinct genetic architecture among related sociological traits.