In silico quality assessment of SNPs—A case study on the Axiom® Wheat genotyping arrays

Abstract

Genotyping arrays proved to be an exemplary tool for the simultaneous analysis of a multitude of single nucleotide polymorphisms (SNPs), a special case of genomic variants. By the example of SNPs represented on the Axiom® Wheat HD genotyping array as well as on the Axiom® Wheat Breeder's genotyping array, we applied a three way classification system to assess the quality of SNPs in bread wheat (Triticum aestivum L.) and subsequently the quality of these genotyping arrays. Class 1 SNPs could be aligned uniquely to the reference genome and did not show any genomic variants in their flanking sequence. Class 2 SNPs could also be aligned uniquely to the reference genome but showed genomic variants in their flanking sequence. The remaining SNPs were assigned to class 3. To determine the number of genomic variants in a SNP's flanking sequence, we used all currently available SNPs in the Ensembl Plants database. From the 819,571 SNPs on the Axiom® Wheat HD genotyping array, we assigned 24,343 to class 1 and from the 35,143 SNPs on the Axiom® Wheat Breeder's genotyping array we classified 2295 SNPs as class 1. We show that class 1 SNPs of the Axiom® Wheat HD genotyping array result in an equidistant coverage of the reference genome. We make the classification table as well as R-scripts available to give breeders and researchers the possibility to reproduce our analysis in an easy way. Moreover, we discuss the possibilities and limitations of such an in silico analysis of genotyping arrays as well as future research possibilities for this approach. Keywords: Triticum aestivum, Single nucleotide polymorphism, SNP classification, Affymetrix, Wheat Breeder's, Wheat HD