Validation of Genotyping by Sequencing Using Transcriptomics for Diversity and Application of Genomic Selection in Tetraploid Potato

Abstract

Potato is an important food crop due to its increasing consumption, and as a result, there is demand for varieties with improved production. However, the current status of breeding for improved varieties is a long process which relies heavily on phenotypic evaluation and dated molecular techniques and has little emphasis on modern genotyping approaches. Evaluation and selection before a cultivar is commercialized typically takes 10-15 years. Molecular markers have been developed for disease and pest resistance, resulting in initial marker-assisted selection in breeding. This study has evaluated and implemented a high-throughput transcriptome sequencing method for dense marker discovery in potato for the application of genomic selection. An Australian relevant collection of commercial cultivars was selected, and identification and distribution of high quality SNPs were examined using standard bioinformatic pipelines and a custom approach for the prediction of allelic dosage. As a result, a large number of SNP markers were identified and filtered to generate a high-quality subset that was then combined with historic phenotypic data to assess the approach for genomic selection. Genomic selection potential was predicted for highly heritable traits and the approach demonstrated advantages over the previously used technologies in terms of markers identified as well as costs incurred. The high-quality SNP list also provided acceptable genome coverage which demonstrates its applicability for much larger future studies. This SNP list was also annotated to provide an indication of function and will serve as a resource for the community in future studies. Genome wide marker tools will provide significant benefits for potato breeding efforts and the application of genomic selection will greatly enhance genetic progress.

Keywords: GATK; SNP annotation; SnpEff; Solanum tuberosum; autotetraploid; crisp score; dry matter; variant discovery.

Figure 1

Heatmap of SNP density across the potato genome. SNP density is shown per chromosome in 100 KB blocks. The color scale indicates the density of markers in that segment of the chromosome (blue, low density; red, high density). SNP density is shown to increase toward the telomeric ends of the chromosomes where gene density is higher. Dark blue regions are indicative of centromeric regions.

Figure 2

Annotation of 183,848 high-quality transcript SNPs using SnpEff showing the classification of SNPs based on effects. The category titled “Other” is comprised of the classes: non-coding transcript exon variant, splice acceptor variant, splice donor variant, splice region variant, start lost, stop gained, and stop lost.

Figure 3

Unrooted neighbor joining dendrogram of the genetic dissimilarity between all samples. Color coding indicates usage type (light blue – French fry; red – crisping; dark blue – fresh).

Figure 4

Manhattan plots of squared marker effects estimated for eye depth (A) and maturity (B) using Bayes A.