Analysis of coding variants in the human FTO gene from the ExAC (gnomAD) Database

Abstract

Single nucleotide polymorphisms (SNPs) in the first intron of the FTO gene (alpha-ketoglutarate-dependent dioxygenase) identified by a genome-wide association study (GWAS) in 2007 continue to be the known variants with the greatest effect on adiposity in different human populations. Currently available data reveal a total of 61 different intronic SNPs associated with adiposity. Coding variants in the FTO gene, on the other hand, have been little explored, but data from complete sequencing of the exomes of various populations are available in public databases and provide an excellent opportunity to investigate potential functional variants in FTO. This study aimed to track nonsynonymous variants in the exons of the FTO gene in different population groups using the ExAC database (gnomAD) (http://exac.broadinstitute.org/) and to analyze the potential functional impact of these variants on the FTO protein. Variants were analyzed using five publicly available pathogenicity prediction programs. Of the 158 mutations identified (152 missense and 6 stop-gain), 64 (40.5%) were classified as pathogenic, 67 (42.4%) were classified as benign, and 27 (17%) were classified as inconclusive. Thirty variants were classified as pathogenic by all five predictors used in this study, and 16 mutations were classified as pathogenic by only one predictor. The largest number of mutations was found in Europeans (non-Finnish) (85/158), all with very low frequencies, and half (32/64) of the variants classified as pathogenic by the five predictors used were also found in this population. The data obtained in this analysis show that a large number of rare coding variants classified as pathogenic or potentially pathogenic by different in silico pathogenicity prediction programs are not detected by GWAS due to the low linkage disequilibrium as well as the limitations of GWAS in capturing rare variants present in less than 1.0% of the population.