Data interpretation: deciphering the biological function of Type 2 diabetes associated risk loci.

ABSTRACT

AIMS: Type 2 diabetes (T2D) is a complex multifactorial disorder with more than 40 loci associated with disease susceptibility. Most of these genome-wide significant loci reside in noncoding regions, it is important to decipher the potential regulatory function of these variants and to differentiate between true and tag signals. Nowadays, databases are being developed to study and predict the function of these associated variants, and RegulomeDB is one such database. METHODS: We used RegulomeDB to analyze the potential function of the associated variants reported in five genome-wide association studies (GWAS) of T2D. RESULTS: We investigated the 1,567 single nucleotide polymorphisms (SNPs) with 989 SNPs with a score of 1-6. Of those 989 SNPs, only 64 returned with RegulomeDB score <3 (evidence of regulatory function), and only four of these were GWAS significant SNPs (THADA/rs10203174, score = 1b; UBE2E2/rs7612463, score = 2a; ARAP1/rs1552224 and TP53INP1/rs8996852, score = 2b). But only 63 % of the annotated SNPs showed regulatory function that is an important limitation of the RegulomeDB as this database only provides information of few regulatory elements. CONCLUSION: This study further supports that some of the noncoding GWAS variants are the true associations and not the tag ones. This study also proves the utility and importance of the RegulomeDB and other such databases. Although it is an extensive database of regulatory elements but has certain limitation due to utilization of only few types of regulatory elements and pathways.