Prediction of functional noncoding variants for autoimmune diseases by machine learning

Abstract

Autoimmune diseases are chronic and intractable. We rarely know radical treatments of those diseases. Genetic factor certainly has its role according to the family history and twin researches. Because multiple genetic factors are associated with a disease simultaneously, we frequently use genome-wide association study (GWAS). However, most GWAS loci are non-coding variants so that we cannot translate their roles, and we cannot pick one locus that is functional because of linkage disequilibrium. In this research, I will introduce how to find functional loci with epigenomic data by machine learning.