Predicting haplotype carriers from SNP genotypes in Bos taurus through linear discriminant analysis

Abstract

BackgroundSNP (single nucleotide polymorphisms) genotype data are increasingly available in cattle populations and, among other things, can be used to predict carriers of specific haplotypes. It is therefore convenient to have a practical statistical method for the accurate classification of individuals into carriers and non-carriers. In this paper, we present a procedure combining variable selection (i.e. the selection of predictive SNPs) and linear discriminant analysis for the identification of carriers of a haplotype on BTA19 (Bos taurus autosome 19) known to be associated with reduced cow fertility. A population of 3645 Brown Swiss cows and bulls genotyped with the 54K SNP-chip was available for the analysis.ResultsThe overall error rate for the prediction of haplotype carriers was on average very low (∼≤1%). The error rate was found to depend on the number of SNPs in the model and their density around the region of the haplotype on BTA19. The minimum set of SNPs to still achieve accurate predictions was 5, with a total test error rate of 1.59.ConclusionsThe paper describes a procedure to accurately identify haplotype carriers from SNP genotypes in cattle populations. Very few misclassifications were observed, which indicates that this is a very reliable approach for potential applications in cattle breeding.

Use of SNP genotypes to identify carriers of harmful recessive mutations in cattle populations

“…These results are in line with those from similar studies (e.g. test error rate ≤1 % for BH2 haplotype in cattle [21], 0−5 % for HLA alleles in humans [24], ∼6 % for casein alleles in cattle [19]), and confirm that this is a highly effective approach with accuracy potentially close to 100 % (virtually faultless) for practical applications in animal genetics.…”

“…The results reported here and -previously- by Biffani et al [21] show that resampling methods offer a valid alternative or complement to GWAS studies, as discussed by Biscarini et al [47]. GWAS studies are in fact known to suffer from some limitations, like susceptibility to spurious associations and poor reproducibility of results [48, 49].…”

“…In a previous publication [21], SNP genotypes were used to predict carriers of the BH2 haplotype in cattle populations. The BH2 haplotype has been associated to stillbirth and peri-natal mortility in calves.…”

“…In the case of SNP-based models, the most predictive SNPs are likely to be linked to the target QTL/mutation. This approach was already presented by Biffani et al [21] to track the position of the BH2 haplotype. The BH2 haplotype is located at 10.140−−11.049 Mb on BTA19 [22], and could be accurately identified through the resampling approach.…”

“…they are often readily available from imputation software like “Beagle” [20]). In a previous study the successful identification of carriers of the BH2 haplotype from SNP genotypes only was described [21]. …”