A low-coverage skim-sequencing and imputation pipeline for genomic selection

Area of Science:

• Plant breeding and genetics
• Genomics
• Bioinformatics

Background:

• Genomic selection (GS) accelerates plant breeding by improving selection accuracy and reducing cycle times.
• Perennial crops like intermediate wheatgrass (IWG) benefit significantly from GS due to long evaluation periods.
• Implementing GS requires affordable, high-density genetic marker systems scalable for large breeding programs.

Purpose of the Study:

• To evaluate the feasibility of using low-coverage whole genome skim-sequencing (skim-seq) for GS in intermediate wheatgrass (IWG).
• To optimize imputation parameters for skim-seq data using STITCH software.
• To compare the predictive accuracy of skim-seq with genotyping-by-sequencing (GBS) for GS in IWG.

Main Methods:

• Implemented ultra-low coverage (0.01x-0.05x) whole genome skim-sequencing for GS at breeding program scale.
• Utilized STITCH (Sequencing to Imputation Through Constructing Haplotypes) software to impute genetic markers.
• Optimized imputation parameters including sequence coverage and ancestral haplotype number.
• Performed cross-validation to compare GS accuracies between skim-seq and GBS data for five traits in IWG.

Main Results:

• Skim-seq data achieved cross-validation accuracies (r = 0.29-0.61) comparable to GBS (r = 0.29-0.55) for five traits in IWG.
• Low-coverage skim-seq, when imputed, provides a viable alternative to GBS for GS in large-genome, polyploid perennial species.
• The developed methods are scalable and applicable to various crops, facilitating GS implementation.

Conclusions:

• Low-coverage skim-seq is a cost-effective and accurate method for genomic selection in intermediate wheatgrass.
• This approach generates archival sequence data robust to technological advancements.
• The scalable methods enable wider adoption of GS in breeding programs for diverse crop species.