Optimizing Selection of the Reference Population for Genotype Imputation From Array to Sequence Variants.

Area of Science:

• Animal Genetics
• Bioinformatics
• Genomic Selection

Background:

• Imputation of high-density genotypes to whole-genome sequences (WGS) is crucial for genomic selection and variant discovery.
• Imputation accuracy depends on reference population size, composition, and allele frequencies.
• Understanding selection strategies for reference populations is key for accurate imputation, especially for common or rare variants.

Purpose of the Study:

• To present and evaluate novel methods for selecting animals for sequencing to optimize reference populations.
• To compare the imputation accuracy achieved using these new methods against existing ones.

Main Methods:

• Simulated WGS data for a dairy cattle population, mirroring real-world linkage disequilibrium and variant frequencies.
• Created reference populations of varying sizes using two novel selection methods (Genetic Diversity Index, Highly Segregating Haplotype) and two established methods.
• Compared imputation accuracies across different reference population strategies.

Main Results:

• The novel methods achieved overall imputation accuracies exceeding 0.85.
• Accuracies for imputing rare variants surpassed 0.50.
• Both novel methods demonstrated superior performance in optimizing reference populations for imputation.

Conclusions:

• For discovering novel trait-associated variants, select animals with unique information using the Genetic Diversity Index method.
• For imputing the entire genotyped population, use the Highly Segregating Haplotype method to select common haplotype carriers.
• These optimized selection strategies enhance the utility of imputed WGS data in cattle breeding and research.