Strategies for genotype imputation in composite beef cattle

Area of Science:

• Animal Genetics
• Bioinformatics
• Genomic Selection

Background:

• Genotype imputation enhances genomic information and reduces costs in cattle breeding.
• Crossbreeding in Brazilian beef cattle alters linkage disequilibrium, posing challenges for accurate genotype imputation.
• Accurate imputation is crucial for genomic evaluations in crossbred populations.

Purpose of the Study:

• To evaluate the optimal genotype imputation strategy for MA genetic group and Canchim cattle.
• To compare the accuracy of various SNP panels for imputing genotypes to the Illumina-BovineHD panel.
• To assess the performance of FImpute and BEAGLE software for genotype imputation.

Main Methods:

• Utilized a dataset of 400 genotyped Canchim and MA cattle.
• Investigated imputation accuracy from lower-density SNP panels (3K to 80K) to the Illumina-BovineHD panel.
• Tested seven reference and target population scenarios, including groupings by birth year, genetic group, and gender.
• Employed FImpute and BEAGLE software, recording computation run-time and measuring accuracy via concordance rate (CR) and allelic R square (R²).

Main Results:

• The highest imputation accuracy was achieved with a reference population of both males and females and a target population of young females.
• SNP panels with at least 50,000 markers (50K, GGP75Ki, GGP80K) were most adequate for imputation to the HD panel in Canchim cattle.
• FImpute significantly reduced computation time (20-100 times faster than BEAGLE) and showed higher efficiency, especially with lower-density panels.

Conclusions:

• Genotyping panels with ≥50,000 markers are suitable for accurate genotype imputation to the HD panel.
• FImpute is an efficient algorithm for genotype imputation, particularly beneficial for lower-density panels.
• These findings support increased genotypic information, reduced costs, and improved genomic selection in crossbred cattle.