Molecular breeding methods such as genomic selection and genome-wide association studies often require high-density genotypic data from many samples, but the cost and complexity of genotyping at this scale may be prohibitive.
This presentation will outline three areas for improvement that can be combined for greatest impact: marker optimization, efficient genotyping, and data imputation.
The challenge with marker optimization is to control cost by minimizing the number of SNPs genotyped yet obtaining an accurate description of each individual genome analyzed and capture the diversity within the breeding germplasm. We will offer methods for minimizing genotyping, allowing the analysis of more individuals, and data imputation generating high density data from sparse genotyping.
The development of a 1K SNP set for genomic selection in soy is presented as an example. The 1,000 soy SNPs were selected to provide maximum informativeness in US North Central Soy public breeding programs. The SNPs were developed into molecular inversion probes, a low-cost genotyping-by-sequencing method. This approach is cost-effective and provides high-quality genotypic data.
This presentation will next discuss PlexSeq, a novel approach for efficient mid-density SNP genotyping. PlexSeq uses artificial intelligence algorithms to create highly multiplexed genotyping assays followed by a unique and effective workflow. This technology has advantages in terms of efficiency and cost.
Next, the webinar will address SNPer, a method of SNP optimization and data imputation in three steps: 1) analysis of the breeding germplasm sequence and haplotype diversity; 2) optimization of SNP targets for genotyping by minimizing the number of targets and maximizing the information gained from each; and 3) imputation of data at ungenotyped SNP loci. The result is a complete and accurate description of each sample genome.