|Sargolzaei, M -|
|Winters, M -|
|Van Kaam, J.B.C.H.M. -|
|Van Doormaal, B -|
|Faust, M -|
|Doak, G -|
Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: September 7, 2012
Publication Date: January 1, 2013
Citation: Van Raden, P.M., Null, D.J., Sargolzaei, M., Wiggans, G.R., Tooker, M.E., Cole, J.B., Sonstegard, T.S., Connor, E.E., Winters, M., Van Kaam, J., Van Doormaal, B.J., Faust, M.A., Doak, G.A. 2013. Genomic imputation and evaluation using high density Holstein genotypes. Journal of Dairy Science. 96(1):668-678. Interpretive Summary: Chips with different densities of single nucleotide polymorphism markers are now available, including a standard with 54,609, higher density with 776,402, and lower densities including 8,653, 6,909 and 2,900 markers. Four evaluation studies were conducted using these five densities of actual genotypes for 161,341 Holsteins. Markers were edited for allele frequency, missing genotypes, parent-progeny conflicts, redundancy, and incorrect genetic map locations to improve imputation. Higher density genotypes provide more markers, but missing alleles must be imputed for animals genotyped at less than highest density. Average reliability of genomic predictions increased only 0.4% with higher density instead of 54,609 markers.
Technical Abstract: Genomic evaluations for 161,341 Holsteins were computed using 311,725 of the 777,962 markers on the Illumina high-density (HD) chip. Initial edits with 1,741 HD genotypes from 5 breeds revealed that 636,967 markers were usable but that half were redundant. Usable Holstein genotypes included 1,510 animals with HD, 82,358 animals with 45,187 (50K) markers, 1,797 animals with 8,031 (8K), 20,177 animals with 6,836 (6K) markers, 52,270 animals with 2,683 (3K) markers, and 3,229 nongenotyped dams (0K) with > 90% of haplotypes imputable because they had 4 or more genotyped progeny. The HD genotypes were from 1,142 American, Canadian, British, and Italian sires, 196 other sires, 138 cows in the USDA Beltsville research herd, and 34 other females. Percentages of correctly imputed genotypes were tested by applying programs findhap and FImpute to an example simulated chromosome with similar population structure but only 1,112 animals with HD. For each chip, 1% of genotypes were missing and 0.02% incorrect initially, and across all animals 89.9% of genotypes were missing initially. After imputation of missing markers with findhap, 99.9% of genotypes were correct from HD, 99.0% from 50K, 94.6% from 6K, 90.5% from 3K, and 93.5% from 0K genotypes. With FImpute, 99.96% were correct from HD, 99.3% from 50K, 94.7% from 6K, 91.1% from 3K, and 95.1% from 0K genotypes. Accuracy for 3K and 6K genotypes further improved about 2% if imputed first to 50K and then to HD instead of all together. Evaluations were tested using imputed actual genotypes and August 2008 phenotypes to predict deregressed evaluations of U.S. bulls proven after August 2008. For 28 traits tested, estimated genomic reliability averaged 61.1% using 311,725 markers vs. 60.7% using 45,187 markers vs. 29.6% from traditional parent average. Squared correlations with future data were slightly greater for 16 traits and slightly less for 12 with HD than with 50K evaluations. The observed 0.4% average increase in reliability was less favorable than the 0.9% expected from simulation but similar to actual gains from other HD studies. The largest marker effects were located at very similar positions, but new markers from the HD chip often had larger effects than the best markers from the 50K chip. Multi-breed evaluation may produce larger gains than the single-breed evaluation investigated here. Increasing the number of HD genotypes above 1,074 did not improve the reliability of Holstein genomic evaluations.