054
Genomic evaluation using QTL information

Monday, August 18, 2014: 11:00 AM
Stanley Park Ballroom (The Westin Bayshore)
Vincent Ducrocq , INRA, UMR1313 GABI, Jouy-en-Josas, France
Pascal Croiseau , INRA, UMR1313 GABI, Jouy-en-Josas, France
Aurélia Baur , INRA, UMR1313 GABI, Jouy-en-Josas, France
Romain Saintilan , INRA UMR 1313 GABI, Jouy-en-Josas, France
Sebastien Fritz , INRA, UMR1313 GABI, Jouy-en-Josas, France
Didier Boichard , INRA, UMR1313 GABI, Jouy-en-Josas, France
Abstract Text:

The French dairy cattle genomic evaluation uses an extension of Marker-Assisted BLUP (MA-BLUP) based on a haplotype model with a residual polygenic effect. A two-step approach to select SNP and construct haplotypes linked to QTL was tested by cross-validation for two breeds with moderate (Montbéliarde) and large (Holstein) reference populations. The comparison with other genomic evaluation approaches showed MA-BLUP validation performances which are very similar for Montbéliarde and slightly lower for Holstein. Some simplifying or ad-hoc assumptions explaining these results have been identified and reveal directions for improvement. In particular, a genome-wide strategy to pinpoint the most informative haplotypes linked to QTL is highly desirable. The robustness and computing simplicity of MA-BLUP remain two appealing features for routine genomic evaluations. The use of haplotypes is expected to give more robust and reliable predictions over more distantly related animals.

Keywords:

dairy cattle

genomic evaluation

genomic selection

quantitative trait locus