|
Abstract:
|
A major goal of human genetics studies of complex disease is to develop genetic risk prediction models and accurate models will have great impacts on disease preventions and treatments. However, despite the identifications of thousands of disease-associated genetic variants through genome wide association studies (GWAS), prediction accuracy remains moderate for most diseases, which is largely due to the challenges in both identifying all the disease-associated variants and accurately estimating their effect sizes. Recent advancements in integrative genomic functional annotation, coupled with the rich collection of summary statistics from GWAS, have enabled increase of statistical power. In this presentation, we introduce GenoPred, the first principled framework incorporating diverse types of annotation data. GenoPred takes GWAS summary statistics as the training data, and adaptively estimates variant effect sizes through modeling of linkage disequilibrium and functional annotations. Compared with state-of-the-art methods based on GWAS summary statistics, GenoPred achieves consistent improvement in prediction performance on on breast cancer, type-II diabetes and six other traits.
|
