Combined Use of Univariate and Multivariate Approaches to Detect Selection Signatures Associated with Milk or Meat Production in Cattle.

OBJECTIVES

The aim of this study was to investigate the genomic structure of the cattle breeds selected for meat and milk production and to identify selection signatures between them.

METHODS

A total of 391 animals genotyped at 41,258 SNPs and belonging to nine breeds were considered: Angus (N = 62), Charolais (46), Hereford (31), Limousin (44), and Piedmontese (24), clustered in the Meat group, and Brown Swiss (42), Holstein (63), Jersey (49), and Montbéliarde (30), clustered in the Milk group. The population stratification was analyzed by principal component analysis (PCA), whereas selection signatures were identified by univariate (Wright fixation index, F) and multivariate (canonical discriminant analysis, CDA) approaches. Markers with F values larger than three standard deviations from the chromosomal mean were considered interesting. Attention was focused on markers selected by both techniques.

RESULTS

A total of 10 SNPs located on seven different chromosomes (7, 10, 14, 16, 17, 18, and 24) were identified. Close to these SNPs (±250 kb), 165 QTL and 51 genes were found. The QTL were grouped in 45 different terms, of which three were significant (Bonferroni correction < 0.05): milk fat content, tenderness score, and length of productive life. Moreover, genes mainly associated with milk production, immunity and environmental adaptation, and reproduction were mapped close to the common SNPs.

CONCLUSIONS

The results of the present study suggest that the combined use of univariate and multivariate approaches can help to better identify selection signatures due to directional selection.