|BAJGAIN, PRABIN - University Of Minnesota
|Rouse, Matthew - Matt
|ANDERSON, JAMES - University Of Minnesota
Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/15/2016
Publication Date: 11/25/2016
Citation: Bajgain, P., Rouse, M.N., Anderson, J. 2016. Comparing genotyping-by-sequencing and Single Nucleotide Polymorphism chip genotyping in Quantitive Trait Loci mapping in wheat. Crop Science. 56:1-17.
Interpretive Summary: Wheat stem rust is a fungal disease of wheat that decreases yield. A strain of the wheat stem rust fungus known as Ug99 emerged in Uganda in 1999 and threatens global wheat production because it is able to infect nearly all wheat varieties. In order to find identify quantitative trait loci (QTL) associated with resistance to Ug99, a biparental population of wheat was assessed in Ethiopia and Kenya. The recombinant inbred lines in each population were genotyped by both a custom chip and sequencing to identify single nucleotide polymorphsims (SNPs). QTL mapping results were compared using the chip data and sequencing data separately. One QTL associated with resistance to Ug99 was detected by both genotyping methods. This study was important because of the resistance identified and also the comparison between genotyping-by-sequencing and genotyping using a custom chip. Overall the two genotyping methods were similar except genotyping-by-sequencing was cheaper and required more computational skills and resources. The resistance-liniked markers identified, once validated, can be utlized in wheat breeding for Ug99 resistant wheat varieties. Ug99 resistant wheat cultivars will protect United States wheat production from yield loss if a Ug99 epidemic were to occur in the United States.
Technical Abstract: Array- or chip-based single nucleotide polymorphism (SNP) markers are widely used in genomic studies because of their abundance in a genome and cost less per data point compared to older marker technologies. Genotyping by sequencing (GBS), a relatively newer approach of genotyping, suggests equal or more appeal because of its even lower cost per data point and due to the avoidance of ascertainment bias during genotyping. In this study, we compared the results from quantitative trait loci (QTL) mapping, marker distribution on linkage maps, genome size, recombination sites covered by the markers, and cost per polymorphic marker, as well as the methodology and workflow between the Illumina Infinium 9,000 SNP-chip genotyping with GBS. Results indicate that while GBS offers similar genome coverage at almost 1/4th the cost of SNP-chip, SNP-chip method is less demanding of computational skills and resources. Eight and nine QTL were detected in the GBS and SNP-chip datasets, respectively, with one QTL common between the systems. Additionally, imputation accuracy of the GBS dataset was examined by introducing missing values randomly and imputing the missing alleles using a probabilistic principal components algorithm. Imputation results suggest recovery of the missing alleles with reasonable accuracy in datasets with low (up to 40%) amount of missing data is possible and can provide acceptable accuracy in gene mapping. Overall, the comparative results indicate that the GBS approach provides good genome coverage and similar mapping results at a much lower cost compared to the SNP-chip genotyping method.