Fig. 3

Effects of the 34 validated haplotypes on female survival. a Results of principal component analysis (PCA) for 34 haplotypes × 12 parameters (Additional file 1: Table S10), selected to reflect the dynamics of increased mortality and slaughter in at-risk versus control mating. These parameters are the days on which 25, 50, 75, and 100% of the maximum deviations were reached after calculating the daily difference in proportions between at-risk (1) and control (0) mating for female cattle that died of natural causes (D), were slaughtered (S), or were still alive (A) over 6 years (Additional file 1: Tables S8, S9). The proportions of variance explained by the principal components are given in parentheses for the first two dimensions. b Results of hierarchical clustering (HC) on principal components. c–h Dynamics of increased mortality and slaughter in at-risk versus control mating expressed as daily difference in proportions between at-risk (1) and control (0) mating for animals that died of natural causes (D), were slaughtered (S), or were still alive (A) over 6 years (Additional file 1: Tables S8, S9). Results are shown for two haplotypes for each of the three PCA-HC clusters (see color code). Note that they include H11P78, which corresponds to the CDH locus (see below), and three other haplotypes that were subjected to clinical and functional investigations later in the article (H5P25, M6P72, and N17P57)