Skip to main content
ARS Home » Research » Publications at this Location » Publication #227880
Title: Genomic Data and Cooperation Result in Faster Progress

Author
item Vanraden, Paul
item Van Tassell, Curtis - Curt
item Wiggans, George
item Sonstegard, Tad
item SCHNABEL, R - UNV OF MISSOURI
item TAYLOR, J - UNIV OF MISSOURI
item SCHENKEL, F - UNIV OF GUELPH

Submitted to: International Committee on Animal Recording(ICAR)
Publication Type: Proceedings
Publication Acceptance Date: 6/18/2008
Publication Date: 1/12/2009
Citation: Van Raden, P.M., Van Tassell, C.P., Wiggans, G.R., Sonstegard, T.S., Schnabel, R.D., Taylor, J.F., Schenkel, F. 2009. Genomic Data and Cooperation Result in Faster Progress. International Committee on Animal Recording (ICAR). ICAR Tech Ser. 13:341–346. 2009.

Interpretive Summary: Traditionally dairy cattle breeders combined information from records and pedigrees when selecting animals. When choosing young animals, often only the parent average was available. With new data from thousands of small DNA segments called single nucleotide polymorphisms (SNP), breeders can now also use this genomic information to get more accurate and reliable estimates of breeding value. This study examined methods to improve the reliability of genomic predictions. Genomic predictions were tested using 2003 historical data for 3,576 bulls born before 1999 to predict current data for 1,759 bulls born 1999-2002. Using more predictor bulls increased the genomic selection reliability over parent average reliability for younger bulls from 4% when 1,151 older bulls were used to 17% with 3,576 older bulls. Genomic predictions of 9 traits comparing 10,000, 20,000 and 40,000 SNP with parent averages showed that all had higher reliabilities than parent averages, with 40,000 SNP having slightly higher reliabilities than the lower SNP densities for all traits. Genomic predictions will greatly enhance the information available to dairy producers in selecting young animals.

Technical Abstract: Genotypes for 38,416 markers of 5,335 Holstein bulls were combined with traditional evaluations to test predictive ability. Genomic evaluations were significantly (P < .0001) more accurate than official parent averages for all 27 traits tested. Squared correlations with future daughter deviations averaged 37% for young bulls as compared to 19% for parent averages. Correlations with future evaluations also increased for older, proven bulls. Genomic gains increased strongly with numbers of bulls genotyped and moderately with numbers of markers. Further increases in reliability are easily possible, for example by obtaining more genotypes for domestic or foreign proven bulls.