Implementation of high-density SNP data for genetic evaluation and QTL discovery in beef cattle

Abstract

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI AT AUTHOR'S REQUEST.] A genomic relationship matrix was built using 1,707 Angus AI sires with complete pedigree information and 698 Angus steers with only paternity information. The genomic matrix was used to estimate breeding values (EBVs) on both sires and steers and residuals were estimated on the steers. EBV and residuals were combined to produce adjusted phenotypes corrected for fixed effects (year and season of birth within feeding pen) for three feed efficiency related traits (average daily feed intake, AFI; residual feed intake, RFI; and average daily gain, ADG). These dependent variables were used for whole genome association analysis. A genome-wide significance of threshold equivalent to p [less than] 0.05 was determined via dependant variable permutation analysis to control experiment wide error rate. Final models for the three traits were assembled using EBVs of steers and explained 54.097%, 62.687% and 54.724% of the genetic variance in AFI, RFI and ADG, respectively. Utility of predicted feed intakes for genetic selection was also tested. Predicted feed intakes (pDMR) were estimated using the Cornell Value Discovery System (CVDS) and were adjusted for fixed effects and EBVs were estimated. Whole genome analysis was performed and concordant SNP associations for ADG, RFI and AFI were identified. Index selection may include pDMR to increase information on feed intake, however correlated responses in ADG are likely if not used within an index.