TVAR: Assessing Tissue-specific Functional Effects of Non-coding Variants with Deep Learning

Abstract Analysis of whole genome-sequencing (WGS) for genetics of disease is still a challenge due to lack of accurate functional annotation of noncoding variants, especially the rare ones. As eQTLs have been extensively implicated in genetics of human diseases, we hypothesize that noncoding rare variants discovered in WGS play a regulatory role in predisposing disease risk. With thousands of tissue- and cell type-specific epigenomic features, we propose TVAR, a multi-label learning based deep neural network that predicts the functionality of noncoding variants in the genome based on eQTLs across 49 human tissues in GTEx. TVAR learns the relationships between high-dimensional epigenomics and eQTLs across tissues, taking the correlation among tissues into account to learn shared and tissue-specific eQTL effects. As a result, TVAR outputs tissue-specific annotations, with an average of 0.77 across these tissues. We evaluate TVAR’s performance on four complex diseases (coronary artery disease, breast cancer, Type 2 diabetes, and Schizophrenia), using TVAR’s tissue-specific annotations, and observe its superior performance in predicting functional variants for both common and rare variants, compared to five existing state-of-the-art tools. We further evaluate TVAR’s G-score, a scoring scheme across all tissues, on ClinVar, fine-mapped GWAS loci, Massive Parallel Reporter Assay (MPRA) validated variants, and observe consistently better performance of TVAR compared to other competing tools.