Genetic diversity analysis of yellow mustard (Sinapis alba L.) germplasm based on genotyping by sequencing

Recent advances in next generation sequencing technologies make genotyping by sequencing (GBS) more feasible for molecular characterization of plant germplasm with complex and unsequenced genomes. We used a GBS protocol consisting of Roche 454 pyrosequencing, genomic reduction and advanced bioinformatics tools to analyze genetic diversity of 24 diverse yellow mustard accessions. One and one half 454 pyrosequencing runs generated roughly 1.2 million sequence reads totaling about 392 million nucleotides. Application of the computational pipeline DIAL identified 512 contigs and 828 SNPs. The BLAST algorithm revealed alignments of 214 contigs with the sequences reported in NCBI nr/nt database. Sanger sequencing confirmed 95 % of 41 selected contigs and 94 % of 240 putative SNPs. The 454 scored SNPs were highly imbalanced among assayed samples. Diversity analysis of these SNPs revealed that 26.1 % of the total variation resided among landrace, cultivar and breeding lines and 24.7 % between yellow- and black-seeded germplasm. Cluster analysis showed that the black-seeded accessions were largely clustered together and the breeding lines were grouped with known origin. Computer simulation was performed to assess the impact of 454 SNPs missing and revealed considerable changes in allelic count, bias in detection of genetic structure, and large deviations from the expected genetic-distance matrix. These findings are useful for parental selection consideration in yellow mustard breeding, and our detailed analyses help illustrate the utility of GBS in genetic-diversity analysis of plant germplasm, particularly for genetic-relationship assessment.