Subclinical ketosis is one of the most dominant metabolic disorders in dairy herds during lactation. Cows suffering from ketosis experience elevated ketone body levels in blood and milk, including β-hydroxybutyric acid (BHB), acetone (ACE) and acetoacetic acid. Ketosis causes serious financial losses to dairy cattle breeders and milk producers due to the costs of diagnosis and management as well as animal welfare reasons. Recent years have seen a growing interest in the use of artificial neural networks (ANNs) in various fields of science. ANNs offer a modeling method that enables the mapping of highly complex functional relationships. The purpose of this study was to determine the relationship between milk composition and blood BHB levels associated with subclinical ketosis in dairy cows, using feedforward multilayer perceptron (MLP) artificial neural networks. The results were verified based on the estimated sensitivity and specificity of selected network models, an optimum cut-off point was identified for the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC). The study demonstrated that BHB, ACE and lactose (LAC) levels, as well as the fat-to-protein ratio in milk, were important input variables in the network training process. For the identification of cows at risk of subclinical ketosis, variables such as BHB and ACE levels in milk were of particular relevance, with a sensitivity and specificity of 0.84 and 0.61, respectively. It was found that the back propagation algorithm offers opportunities to integrate artificial intelligence and dairy cattle welfare within a computerized decision support tool.

Taking into account functional traits in the breeding practice should lead to a longer productive life of cows. However, despite the increased contribution of these traits in bull selection indices, their daughters are frequently culled as early as the 2nd or 3rd lactation. The problem is whether and to what extent the genetic potential of animals is realized in the production practice. Therefore, the purpose of this study was to determine the associations between the breeding value (BV) of bulls and their daughters for cow longevity and culling reasons in the Holstein-Friesian cattle population in Poland. Data for 532 062 cows culled in 2012, 2015, and 2018 were analyzed. A majority of 5 045 cow sires originated from Poland, Germany, France, the Netherlands, and the United States. The highest variation in the contribution of culling reasons was for the cows culled at the age of 2-4 years. The contribution of the culling reasons, analyzed in relation to the cow culling age, remained similar and the only exception was culling because of old age, for which a significant increase was observed only for the culling age of at least 9 years (13.8%), which was reached by only 7.3% of the cows. The sires were characterized by generally high BV for conformation and reproductive traits. However, they had, at most, the average genetic potential for functional longevity. There were a number of beneficial associations found between the BV of bulls and the distribution of culling reasons in their daughters. For example, it concerns relations between the somatic cell score in milk and culling due to udder diseases and low milk yield, between the interval from calving to first insemination and low milk yield, between the protein yield and old age, or between the BV for certain conformation traits (size, udder) and cow culling due to age. In these cases, as the BV increased for a given trait, the contribution of the corresponding cow culling reason tended to decrease. Our study showed that it seems reasonable to consider Holstein-Friesian cows aged at least 9 years at culling to be long-living animals. This is primarily evidenced by the rapid increase in the culling due to old age in relation to younger cows. Nowadays the above age limit can be suggested as a criterion of longevity for Holstein-Friesian cows but the criterion should be updated to the relation genotype-environment-economy that tends to change over time.

The serious drawback underlying the biological annotation of whole-genome sequence data is the p >> n problem, which means that the number of polymorphic variants (p) is much larger than the number of available phenotypic records (n). We propose a way to circumvent the problem by combining a LASSO logistic regression with deep learning to classify cows as susceptible or resistant to mastitis, based on single nucleotide polymorphism (SNP) genotypes. Among several architectures, the one with 204,642 SNPs was selected as the best. This architecture was composed of two layers with, respectively, 7 and 46 units per layer implementing respective drop-out rates of 0.210 and 0.358. The classification of the test data resulted in AUC = 0.750, accuracy = 0.650, sensitivity = 0.600, and specificity = 0.700. Significant SNPs were selected based on the SHapley Additive exPlanation (SHAP). As a final result, one GO term related to the biological process and thirteen GO terms related to molecular function were significantly enriched in the gene set that corresponded to the significant SNPs. Our findings revealed that the optimal approach can correctly predict susceptibility or resistance status for approximately 65% of cows. Genes marked by the most significant SNPs are related to the immune response and protein synthesis.

The goal of our study was to identify the SNPs, metabolic pathways (KEGG), and gene ontology (GO) terms significantly associated with calving and workability traits in dairy cattle. We analysed direct (DCE) and maternal (MCE) calving ease, direct (DSB) and maternal (MSB) stillbirth, milking speed (MSP), and temperament (TEM) based on a Holstein-Friesian dairy cattle population consisting of 35,203 individuals. The number of animals, depending on the trait, ranged from 22,301 bulls for TEM to 30,603 for DCE. We estimated the SNP effects (based on 46,216 polymorphisms from Illumina BovineSNP50 BeadChip Version 2) using a multi-SNP mixed model. The SNP positions were mapped to genes and the GO terms/KEGG pathways of the corresponding genes were assigned. The estimation of the GO term/KEGG pathway effects was based on a mixed model using the SNP effects as dependent variables. The number of significant SNPs comprised 59 for DCE, 25 for DSB and MSP, 17 for MCE and MSB, and 7 for TEM. Significant KEGG pathways were found for MSB (2), TEM (2), and MSP (1) and 11 GO terms were significant for MSP, 10 for DCE, 8 for DSB and TEM, 5 for MCE, and 3 for MSB. From the perspective of a better understanding of the genomic background of the phenotypes, traits with low heritabilities suggest that the focus should be moved from single genes to the metabolic pathways or gene ontologies significant for the phenotype.

The objective of this study was to estimate genetic parameters for the analyzed fertility traits of Polish Holstein-Friesian primiparous and multiparous cows, as a step toward the incorporation of new traits into routine genetic evaluation. Lactation records of 116,836 Polish Holstein-Friesian cows were studied. The records cover the first three lactations of all cows. The cows, daughters of 2884 sires, were calved from 2006 to 2020. The conception rate for heifers (CRh) and for cows in the second (CR2) and third parity (CR3), the interval from first calving to first insemination (CTFI), the days open (DO), and the interval from first to successful insemination for heifers (FSh) and for cows in second (FS2) and third (FS3) parity were analyzed. The BLUPf90 package and a Bayesian method via Gibbs sampling were used to estimate (co)variance components. In general, all heritabilities were low and ranged from 0.013 (CTFI) to 0.038 (FS2). The heritability of conception rate and interval from first to successful insemination was slightly lower for heifers than for cows. Genetic correlations were moderate to high with two exceptions: correlation of CTFI with CRh (0.033) and with FSh (−0.051). The results indicate that a few analyzed traits could be used in genetic evaluation of Polish Holstein-Friesian cows. It is suggested to complement the selection index with traits for both heifers and cows, such as the interval from first to successful insemination (i.e., FSh, FS2 and FS3), in order to enable effective improvement of female fertility scores in the Polish Holstein-Friesian population.

Fourier transform infrared spectroscopy (FTIR) allows for the determination of milk acetone (mACE) and beta-hydroxybutyrate (mBHB) concentrations, providing a potential herd monitoring tool for hyperketolactia, defined as elevated milk ketone bodies. The study aim was to characterize mACE and mBHB concentration dynamics during early lactation in Polish Holstein-Friesian cows. Milk samples (n = 3,867,390) were collected within 6 to 60 days in milk (DIM) over a 4-yr period (April 1, 2013 to March 31, 2017) from approximately 21,300 dairy herds (average 38.7 cows/herd). Fixed effects of parity, DIM, and their interaction on mACE and mBHB concentrations were determined using a mixed model with a herd-year-season fixed effect and random cow effect. Published hyperketolactic mACE (>= 0.15 mmol/L) and mBHB (>= 0.10 mmol/L) threshold concentrations were used to classify study milk samples into ketolactia groups of normal (mACE <0.15 mmol/L and mBHB = 0.15 mmol/L or mBHB >= 0.10 mmol/L). Additionally, HYKL samples were categorized into subpopulations as having elevated mBHB and mACE (HYKLACEBHB, mACE >= 0.15 mmol/L and mBHB >= 0.10 mmol/L), only elevated mBHB (HYKLBHB; mACE = 0.10 mmol/L), or only elevated mACE (HYKLACE; mACE >= 0.15 mmol/L and mBHB <0.10 mmol/L). Effects of parity, DIM, ketolactia group or subpopulation, and their interactions on mACE and mBHB concentrations were also determined using the mixed model that included ketolactia group or subpopulation as an independent variable. Across all data, niACH and mBHB concentrations were influenced by effects of parity, DIM, and their interaction as well as parity, DIM, ketolactia group or subpopulation, and their interactions. For all samples, mACE and mBHB concentrations decreased with increasing DIM, with mACE concentration declining more rapidly compared with mBHB. In the data set, 68% and 32% of all samples were defined as normal or HYKL, respectively. Among HYKL samples, mACE was elevated soon after calving and declined over time. In contrast, mBHB started lower after calving and increased reaching peak concentrations around 30 DIM, and then decreased. Within HYKL samples, 50.8, 41.3, and 7.9% were categorized as HYKLACEBHB, HYKLBHB, and HYKL ACE respectively. Between 6 and 21 DIM, 11.3% of HYKL were classified as HYKLACE. Primiparous cows had greater (14.8%) HYKLACE samples in this time period. In conclusion, this study has characterized mACE and mBHB concentrations during early lactation and determined effects of parity, DIM, and their interaction. Using published criteria interpreting mACE and mBHB concentrations, it was intriguing to identify a unique population of samples having elevated mACE without mBHB in early lactation, especially in primiparous cows. Further research is needed to determine if this sample population represents an unhealthy metabolic status that adversely affects cow health and performance.