JS-MA: A Jensen-Shannon Divergence Based Method for Mapping Genome-wide Associations on Multiple Diseases

Abstract

Taking advantages of high-throughput genotyping technology of single nucleotide polymorphism (SNP), large genome-wide association studies (GWASs) have been considered as the promise to unravel the complex relationships between genotypes and phenotypes, in particularly common diseases. However, current multi-locus-based methods are insufficient, in terms of computational cost and discrimination power, to detect statistically significant interactions and they are lacking in the ability of finding diverse genetic effects on multifarious diseases. Especially, multiple statistic tests for high-order epistasis ($ \geq $ 2 SNPs) will raise huge analytical challenges because the computational cost increases exponentially as the growth of the cardinality of SNPs in an epistatic module. In this paper, we develop a simple, fast and powerful method, named JS-MA, using the Jensen-Shannon divergence and a high-dimensional $ k $-mean clustering algorithm for mapping the genome-wide multi-locus epistatic interactions on multiple diseases. Compared with some state-of-the-art association mapping tools, our method is demonstrated to be more powerful and efficient from the experimental results on the systematical simulations. We also applied JS-MA to the GWAS datasets from WTCCC for two common diseases, i.e. Rheumatoid Arthritis and Type 1 Diabetes. JS-MA not only confirms some recently reported biologically meaningful associations but also identifies some novel findings. Therefore, we believe that our method is suitable and efficient for the full-scale analysis of multi-disease-related interactions in the large GWASs.