Statistical aspects of cattle MHC (BoLA) and disease associations exemplified by an investigation of subclinical mastitis

Abstract

A preliminary investigation of associations between cattle MHC (BoLA) and subclinical mastitis measure by somatic cell counts (SCC) of quarter milk samples of 239 animals from 3 different breeds and crossbreeds is presented. the statistical analysis is based on a statistical model which includes effects of external and genetic relationship groups (fixed or random) as well as effects of MHC genotypes or alleles. With this model it is possible to estimate unweighted allelic effects of all alleles simultaneously by multiple regression. An analysis of variation in cow average logarithmic SCC indicated no significant random sire effect, no significant influence of BoLA alleles but a significant effect of herds, lactation numbers and breeds. These results are discussed in relation to other investigations of MHC associations in cattle concerning mastitis, bovine leukosis and tick and worm infestations. A general problem of demonstrating significant and/or consistent disease associations from population data is discussed in relation to loci studied, disease parameters studied, experimental design and selection on MHC.