Comparative Genotyping-by-Sequencing Analyses in a Soft Red Winter Wheat Bi-Parental Mapping Population with Differing Levels of Sequencing Coverage

With the advent and increasing popularity of genotyping-by-sequencing (GBS) as a tool for use in association studies as well as genomics approaches to breeding, one area of interest is the level of coverage and genomic reduction at which one obtains ‘optimal’ amounts of sequence. One of the drawbacks to GBS is that there often exists a large amount of missing data between individuals within a given analysis and also between analyses. Factors such as genome size, level of coverage, cost of sequencing, and the number of desired markers all must be evaluated to determine how much sequence to obtain. We investigated the level of optimal coverage in a soft red winter wheat bi-parental mapping population (‘Clark’ x IL97-1828). We obtained two lanes of sequence from an Illumina HiSeq 2000 run with the same 96 individuals per lane, including the two parents and 94 progeny. We used a two-enzyme restriction digest GBS protocol with PstI and MspI for library preparation and genomic reduction prior to sequencing. Sequence data was analyzed using the TASSEL GBS Java Pipeline. We performed separate analyses for: 1) each lane individually, 2) each lane individually with the parental sequences from the other lane added in (essentially single lanes with 2x parents), and 3) both lanes combined. Here we present a comparison of these analyses including tag counts, SNP counts, and genetic maps created from the different data sets.