Genetic Merit Research Articles

Principal component analysis revisited: fast multitrait genetic evaluations with smooth convergence.

A cornerstone in breeding and population genetics is the genetic evaluation procedure, needed to make important decisions on population management. Multivariate mixed model analysis, in which many traits are considered jointly, utilizes genetic and environmental correlations between traits to improve the accuracy. However, the number of parameters in the multitrait model grows exponentially with the number of traits which reduces its scalability. Here, we suggest using principal component analysis to reduce the dimensions of the response variables, and then using the computed principal components as separate responses in the genetic evaluation analysis. As principal components are orthogonal to each other so that phenotypic covariance is abscent between principal components, a full multivariate analysis can be approximated by separate univariate analyses instead which should speed up computations considerably. We compared the approach to both traditional multivariate analysis and factor analytic approach in terms of computational requirement and rank lists according to predicted genetic merit on two forest tree datasets with 22 and 27 measured traits, respectively. Obtained rank lists of the top 50 individuals were in good agreement. Interestingly, the required computational time of the approach only took a few seconds without convergence issues, unlike the traditional approach which required considerably more time to run (7 and 10 h, respectively). The factor analytic approach took approximately 5-10 min. Our approach can easily handle missing data and can be used with all available linear mixed effect model softwares as it does not require any specific implementation. The approach can help to mitigate difficulties with multitrait genetic analysis in both breeding and wild populations.

Isoacid supplementation influences feed sorting, chewing behaviors, and enteric methane emissions differentially in mid-lactation dairy cows depending on dietary forage level

Past studies have shown that isoacids (ISO) improve dairy cow performance, with effects varying based on dietary forage levels, leading us to speculate that ISO supplementation may also differentially affect enteric methane (CH4) emissions depending on dietary forage levels. Therefore, our primary objective was to examine the effects of ISO supplementation on enteric CH4 emissions in lactating dairy cows fed 2 forage NDF levels (FL), along with monitoring feed particle sorting and chewing behaviors to assess any potential interactions. Sixty-four (64) mid-lactation Holstein cows were utilized in a 10-wk long randomized complete block design trial. Parity, DIM, and prior milk yield (MY) for multiparous cows or genetic merit for primiparous cows were used as blocking factors. Cows were randomly assigned to 1 of 4 diets (n = 16 per diet) with a 2 × 2 factorial arrangement of treatment combinations, including 2 FL, 17 (LF) and 21% forage NDF (HF), without or with ISO supplementation (7.85 mmol/kg DM for isobutyrate and 3.44 mmol/kg DM for 2-methylbutyrate, respectively). Enteric CH4 and chewing activity (rumination and eating time) were measured using the GreenFeed system and sensor-based ear tag system, respectively. The particle size of each diet and ort from individual cows was measured using the Penn State Particle Separator, and a sorting index was calculated. A sorting index of 100 indicates no sorting, while values above or below 100 indicate sorting for or against, respectively. Data were analyzed using a mixed model including FL, ISO, and FL × ISO as fixed effects and block as a random effect (lme4 in R). Our result shows that ISO increased sorting index for long particle in LF (96.1 vs. 109; P < 0.01) but decreased it in HF (100.8 vs. 92.5; P = 0.04). In contrast, ISO did not affect the physically effective particle sorting index (P = 0.51) or intake (P = 0.27) regardless of FL. In alignment with the long particle sorting index, ISO decreased eating and chewing time in the HF but increased them in the LF diet (P < 0.01). In contrast, rumination time was comparable between FL (P = 0.70) and ISO levels (P = 0.19). In the LF diet, ISO supplementation reduced daily CH4 production (g/d) by 9% and intensity (g/kg of MY) by 18% (P < 0.01). In the HF diet, ISO supplementation led to a 10% increase in daily CH4 (P < 0.01) but did not change CH4 intensity (P = 0.17; g/kg of MY) due to improved milk production. Overall, ISO altered feed sorting, feeding behaviors and enteric CH4 emissions depending on FL.

Supplementation of isoacids to lactating dairy cows fed low or high forage diets: Effects on performance, digestibility, and milk fatty acid profile

Our objective was to determine the effects of isoacids (ISO) on the lactation performance, digestibility, and milk fatty acid (FA) profile of Holstein cows fed 2 forage NDF levels (FL). The study was 10-wk long (including 2-wk for covariate) utilizing a randomized complete block design. Sixty-four mid-lactating Holstein cows [662 ± 71 kg BW, 119 ± 51 DIM, 2 ± 0.9 parity] were blocked by parity, DIM, and prior milk yield (MY) for multiparous cows or genetic merit for primiparous cows, and randomly assigned to 1 of the 4 diets (n = 16). Diets were arranged as a 2 × 2 factorial, with 2 FL containing 21 (HF) and 17% forage NDF (LF) without (WIA) or with ISO supplementation (IA, 7.85 mmol/kg DM and 3.44 mmol/kg DM for isobutyrate and 2-methylbutyrate, respectively). Diets were balanced for similar NEL (1.58 Mcal/kg DM), CP (16.0%) and total NDF (27.2%). Feed intake and MY were recorded daily. Nutrient digestibility for each cow was determined using indigestible NDF as a marker, and fecal samples were collected at 8-time points (4 h intervals between samples). Individual cow milk samples composited over a 10-wk period were analyzed using gas chromatography for FA profile. The statistical model included FL, ISO, and FL × ISO as fixed effects and block as a random effect (lme4 in R). The ISO did not affect DMI (P = 0.13), while LF had greater DMI than HF diets (27.8 vs. 26.0 kg/d; P < 0.01). However, ISO increased MY (34.7 vs. 37.2 kg/d; P < 0.01) and ECM (41.9 vs. 39.0 kg/d; P < 0.01) by 7% in HF but not in the LF diet, suggesting FL × ISO interaction (P = 0.04). Interestingly, ISO increased ADG (0.4 kg/d) but decreased MUN by 9% only in LF diet as indicated by FL × ISO interaction (P < 0.01). Additionally, ISO increased DM, OM, NDF, and CP digestibility by 10-24% in HF (P < 0.01), but not in LF (FL × ISO; P > 0.05). As expected, ISO increased milk odd chain FA profiles in the IA groups irrespective of FL, e.g., the IA had greater C15:0 (1.87 vs. 1.54 g/100g FA; P = 0.03) and a tendency to be greater C17:0 levels (0.86 vs. 0.76 g/100g FA; P = 0.05) compared with WIA groups. Overall, ISO improved MY and nutrient digestibility in the HF whereas it increased ADG and decreased MUN in LF diet. Additionally, ISO increased milk odd chain FA (C15:0 and C17:0) regardless of FL.

PSX-4 Exploring the impact of age and body condition score on the genetic analysis of mature cow weight in American Angus cattle

Abstract Selecting sires with optimal genetic merit for mature weight (MWT) is crucial for profitability of beef producers, ensuring cost-effective production and increased economic returns per animal. The main objective of this study was to determine how different statistical approaches [e.g., including the fixed effects (FE) of age and body condition score (BCS) in the statistical model and/or pre-adjusting the weights for these effects] affect variance components estimation and genetic parameters for MWT, and their impact in the posterior selection decisions. The dataset provided by the American Angus Association (AAA) comprised 80,296 MWT records from 37,934 cows aged 2 to 15 yr. Contemporary groups (CG) were defined by concatenating herd, year, and season of measurement. Model 1 analyzed unadjusted MWT with CG and BCS as categorical FE and age as a covariable (quadratic). Model 2 was model 1 without BCS, model 3 was model 1 without age, and model 4 considered CG as the only FE. In model 5, MWT was pre-adjusted for 6 yr of age and BCS of 6, and CG was the only FE included in the statistical model. Additive genetic and permanent environment (pe) random effects were included in all statistical models except model 5, which did not include pe (only one adjusted record per animal). Variance components were calculated using REML implemented in the BLUPf90+ family of programs. Sires with at least 10 offspring (n = 1,483) were ranked according to their breeding values (EBVs) predicted for MWT, and all and the top 10% sires were used to estimate the rank correlation across models. The heritability (± SE) estimated for MWT was 0.43 (± 0.011), 0.42 (± 0.011), 0.70 (± 0.008), 0.72 (± 0.082), and 0.37 (± 0.013) for models 1, 2, 3, 4, and 5, respectively. Repeatability values were 0.64, 0.67, 0.72, and 0.76 for models 1, 2, 3, and 4, respectively. The EBV ranks were highly correlated between models 1 and 2 (Top 10%: 0.96; All: 0.99), and models 3 and 4 (Top 10%: 0.99; All: 0.99). Inflated heritabilities in models 3 and 4 suggest that the effect of age is confounded with the additive genetic effect, which will not translate to greater genetic progress. Therefore, disregarding age in the genetic evaluation of MWT is not recommended. Similarities between models 1 and 2 suggest that BCS has a relatively small impact on MWT, probably related to the relatively low genetic correlation between these traits. Models 1 and 5 cannot be directly compared, as they use slightly different response variables (body weight vs. pre-adjusted weight) and may lead to slightly different selection decisions (correlation between Top 10%: 0.85; All: 0.95). Further investigations in this field are warranted.

313 Many farmers want a prediction of future performance, not necessarily breeding potential

Abstract High heritability is a characteristic of many performance traits in cattle translating to a relatively strong correlation between an (accurate) estimated breeding value for an animal and its (future) production potential. However, the broad sense heritability (i.e., numerator in the formula includes both additive and non-additive genetic variance) is often larger than the narrow sense heritability (i.e., numerator includes only the additive genetic variance) especially for low heritability traits like reproductive performance. Moreover, the repeatability of a trait can be greater again, the difference between the heritability and repeatability is the inclusion of a permanent environmental variance in the numerator. Accurate and pertinent omic-based animal-level features can contribute to an increase in narrow sense heritability owing to a possible reduction in the residual variance; such data can also provide insights into both non-additive genetic variance and the permanent environmental variance. Even at the level of the contemporary group, omic-level data from the likes of bulk blood/saliva/milk sample, can improve how systematic environmental effects are modeled in the mixed model equations thus also reducing the residual variance. The greater heritability will contribute to a strengthening of the correlation between future phenotypic performance potential and estimated additive genetic merit. Importantly, omic measures for individual animals can help predict the permanent environmental effect of that animal capturing the yet untapped contributor to phenotypic variance. Irrespective, predictions need to be replaced by prescriptions; end users are inundated with data yet crave actionable information. Animal breeders have a lush (pardon the pun!) history of distilling estimates of genetic merit for a plethora of traits into a single value per animal (i.e., breeding indexes); this is the first step in transitioning from predictions to prescriptions. Given the familiarity of end users with breeding indexes, the potential exists to develop production indexes to rank commercial animals on future expected performance taking cognizance of genetic and non-genetic effects. Such production indexes align well with the associated breeding index but where the estimates of genetic merit for the traits in question are replaced by estimates of production merit; these estimated production values are the sum of the additive genetic, non-additive genetic and permanent environmental effects. The weighting factors in the production indexes are more short-term using current costs and prices as opposed to breeding indexes which are longer term. In comparison with the breeding indexes, the production indexes can include ancillary information like gender, age, and month of calving/birth. Like breeding indexes, they are updated in real time. Examples of the application of genome-enabled production indexes are guides for valuing animals for purchase or culling.

115 Confidence intervals for validation statistics with data truncation in genomic prediction

Abstract Validation by data truncation is a common practice in genetic evaluations because of the interest in predicting the genetic merit of a set of young selection candidates. Two of the most used validation methods in genetic evaluations are predictivity (correlation between pre-adjusted phenotypes and EBV) divided by the square root of the heritability) and the linear regression (LR) method. Both methods compare predictions with the whole and partial datasets obtained by removing the information related to a set of validation individuals. Confidence intervals (CI) for predictivity and the LR method can be obtained by k-fold validation or bootstrapping. Analytical or frequentist CI are unavailable for predictivity and the LR method and would be beneficial to avoid running several validations. The analytical CI can also help test the quality of bootstrap intervals. This study aimed to derive analytical CI for predictivity, and statistics included in the LR method (bias, dispersion, ratio of accuracies, and reliability). The CI for the bias, dispersion, and reliability depends on the (co)variances of the EBV across the individuals in the validation set. The CI for the ratio of accuracies and predictivity were obtained through the Fisher transformation. We showed the adequacy of the analytical CI using simulation. The analytical CI were closer to the simulated ones. Bootstrap CI tend to be narrower than the simulated ones. Estimating the sampling variation of predictivity and the statistics in the LR method without replication or bootstrap is possible for any dataset with the method proposed in this study. All the formulas derived in this study were implemented in a new program belonging to the BLUPF90 suite called Validationf90.

195 Effect of selection for growth and maternal influence on crossbred Angus heifer development

Abstract This multiyear experiment was designed to determine the effects of sire yearling weight (YW) and maternal influence (MILK) expected progeny differences (EPD) on production and fertility of beef replacement heifers. Sires were classified as moderate (&amp;lt; 50th percentile) or high for YW EPD (122 ± 29; from 79 to 172) and as low (&amp;lt; 50th percentile) or high for MILK EPD (27 ± 8; from 5 to 40). Fall-born (n = 214) and spring-born (n = 207) heifers were enrolled in the study. Birth weight (BW), weaning weight (WW), and YW of the heifers were measured. In addition, reproductive tract score (RTS), dry matter intake (DMI), and average daily gain (ADG) while consuming an unprocessed grass hay diet were determined in a subsample of enrolled heifers. Neogen Igenity Beef scores (NIBS) were also run on this subset. Pregnancy rate and conception rate at fixed timed artificial insemination (FTAI) were also determined. Adjusted BW, WW and YW and RTS were analyzed using the GLIMMIX procedure of SAS (9.4, SAS Institute Inc., Cary, NC), with YW and MILK EPD as fixed effects and year of birth as random effect. Effects of sire YW and MILK classification on the likelihood of pregnancy and successful FTAI were also analyzed using the GLIMMIX procedure of SAS. Pearson correlation coefficients between independent and dependent variables were obtained using the CORR procedure of SAS. Sire MILK EPD had no effect (P ≥ 0.13) on the body weight of fall- or spring-born heifers. However, fall-born heifers were 0.12 kg and 0.34 kg heavier (P = 0.06 and 0.001, respectively) per unit increase in sire YW EPD at weaning and yearling time, respectively. There was no effect of sire YW EPD on fall-born heifer DMI as percentage of body weight (DMI % BOW; P = 0.74) nor ADG (P = 0.90) while consuming the low-energy diet. Spring-born heifers had significantly greater (all, P ≤ 0.001) adjusted BW, WW, and YW (0.02, 0.23 and 0.37 kg per unit increase in sire YW EPD, respectively. Reproductive tract score was strong and negatively correlated (P ≤ 0.001) with DMI % BOW. Heifer YW NIBS were moderate and positively correlated (P &amp;lt; 0.001) with ADG. There was a tendency (P = 0.09) for a weak and negative correlation between residual feed intake genetic scores and DMI % BOW. Sire genetic merit for yearling growth was transmitted to replacement heifers. However, neither ADG, nor unprocessed forage DMI % BOW was influenced by genetic potential for growth during this study. Further research is needed to assess the benefits of using genomic profile of heifers for maternal, performance and carcass traits, alongside sire EPDs, when selecting and developing beef replacement heifers.

Lack of a genetic risk continuum between pubertal timing in the general population and idiopathic hypogonadotropic hypogonadism.

Pubertal timing is a highly heritable trait in the general population. Recently, a large-scale exome-wide association study has implicated rare variants in six genes (KDM4C, MC3R, MKRN3, PDE10A, TACR3, and ZNF483) as genetic determinants of pubertal timing within the general population. Two of the genes (TACR3, MKRN3) are already implicated in extreme disorders of pubertal timing. This observation suggests that there may be a pervasive "genetic risk continuum" wherein genes that govern pubertal timing in the general population, by extension, may also be causal for rare Mendelian disorders of pubertal timing. Hence, we hypothesized that the four novel genes linked to pubertal timing in the population will also contribute to idiopathic hypogonadotropic hypogonadism (IHH), a genetic disorder characterized by absent puberty. Exome sequencing data from 1322 unrelated IHH probands were reviewed for rare sequence variants (RSVs) (minor allele frequency bins: <1%; <0.1%; <0.01%) in the six genes linked to puberty in the general population. A gene-based rare variant association testing (RVAT) was performed between the IHH cohort and a reference public genomic sequences repository-the Genome Aggregation Database (gnomAD). As expected, RVAT analysis showed that RSVs in TACR3, a known IHH gene, were significantly enriched in the IHH cohort compared to gnomAD cohort across all three MAF bins. However, RVAT analysis of the remaining five genes failed to show any RSV enrichment in the IHH cohort across all MAF bins. Our findings argue strongly against a pervasive genetic risk continuum between pubertal timing in the general population and extreme pubertal phenotypes. The biologic basis of such distinct genetic architectures' merits further evaluation.

Effects of genetic and environmental trends from 1970 to 2020 on farm efficiency estimated with a whole farm modeling system

The objectives of this study were to evaluate associations of genetic, cow management and nutrition, inbreeding, and crop yields from 1970 to 2020 with measures of production and economic efficiency according to a whole farm model, and to evaluate effects of genetic change in individual traits on economic efficiency in comparison to expectations from economic selection indexes. Genetic and phenotypic performance metrics for Holsteins from 1970 and 2020 were retrieved and input into the Integrated Farm System Model (IFSM) for a 7000 cow TX herd and a 50 cow PA grazing herd. Crop yields estimates from 1970 and 2020 were retrieved and farm hectarage was altered so that forage and energy concentrate requirements were met through farm production; likewise, scenarios evaluating effects of atmospheric CO2 fertilization (CO2F) on crop yield were evaluated by altering farm hectarage. For single traits that could be dynamically modeled by IFSM, performance shifts and resulting change in product prices or management expenses were added to 1970 base models. Economic efficiency was evaluated as the per cow return to management and unpaid factors as compared with 1970 base models. As averaged across state scenarios, gains in economic efficiency were +$945 and -$76 for additive genetic and inbreeding effects, respectively, for a total gain from genetic change of +$869. Genetic gain in fat yield (+$549) and protein yield (+$524) were responsible for most of the genetic gain, whereas milk yield (-$128) and increased cow BW (-$129) depressed economic efficiency. Genetic change in productive life had a smaller effect (+$44) than predicted unless heifers were purchased and at double the default value. Gains due to cow management and nutrition increased efficiency by +$666 and crop yield increased efficiency by +$711, of which +$371 was attributed to CO2F across scenarios. Whole farm dry matter efficiencies derived as the ratio of fat and protein corrected milk yield to dry matter intake increased from 0.82 (PA) and 0.97 (TX) in 1970 to 1.20 in 2020 and could be higher if farms reduce the size of their replacement herd by producing beef calves. The landmass required in 2020 was 63% and 78% of the 1970 requirement for TX and PA, respectively. Changes in cow genetic merit, management and nutrition, and crop yields have all increased the economic and environmental sustainability of milk production, and systems such as IFSM could be a useful tool to help inform economic selection indices.

Mastitis has a cumulative and lasting effect on milk yield and lactose content in dairy cows

Milk lactose content (LC) physiologically decreases with parity order in dairy cows, but also after udder health inflammation(s) and/or in presence of elevated milk SCC in subclinical cases. Therefore, the progressive decrease in milk LC observed along cows' productive life can be attributed to a combination of factors that altogether impair the epithelial integrity, resulting in weaker tight junctions, e.g., physiological aging of epithelium, mechanical epithelial stress due to milking, and experienced clinical or subclinical mastitis. Mastitis is known to affect the udder synthesis ability too, so our intention through this study was to evaluate if there is a cumulative and lasting effect of mammary gland inflammation(s) on milk yield (MY) and LC. For this purpose, we used diagnoses of clinical mastitis and milk data of Austrian Fleckvieh cows to evaluate the effect of cumulative mastitis events on LC and MY. Only mastitis diagnoses recorded by trained veterinarians were used. Finally, we investigated if cumulative mastitis is a heritable trait and whether it is genetically correlated with either LC or MY. Estimates were obtained using univariate and bivariate linear animal models. A significant reduction in LC and MY was observed in cows that suffered from mastitis compared with those that did not experience udder inflammation. The h2 of cumulative mastitis is promising and much greater (0.09) than the h2 of the binary event itself (≤0.03). The genetic correlations between cumulative mastitis with LC and MY were negative, suggesting that cows with a great genetic merit for MY and LC are expected to be more resistant to repeated inflammations and less recidivist. When we used number of lifetime SCC peaks (≥200,000 or 400,000 cells/mL) to calculate cumulative inflammation events, h2 was even higher (up to 0.38), implying that also subclinical mastitis has a relevant negative impact on both LC and MY. Finally, the present study demonstrated how repeated mastitis events can permanently affect the mammary gland epithelial integrity and synthesis ability, and that the number of cumulative mastitis is a promising phenotype to be used in selection index in combination with other indicator traits toward more resistant and resilient mammary glands.

Multivariate analysis of herd structure and genetic resource indicators in seedstock beef cattle herds.

Genetic, environmental, technological and financial resources are used differently in cattle herds that participate in the same breeding programme. The percentages of calves sired by sires within their own herd or from external herds vary across herds, as do the intensities of use of reproductive biotechnologies. These divergences may be related to differences in the indicators of genetic performance for economic traits. The aim of this study was to determine the factors related to herd structure and genetic resource utilization that exert the greatest influence on the genetic merit of seedstock herds within a Nellore breeding programme. The database comprised 21 factors, along with genomic-enhanced expected progeny differences (GE-EPDs) for growth, reproductive and carcass traits, as well as a selection index of animals from 128 herds. By combining principal component analysis and cluster analysis, we were able to group the herds. We identified statistically significant differences (p < 0.05) in the mean values of the factors, GE-EPDs and genetic trends among the groups of herds. Differences in the percentage of sires from external herds and in sire age between the groups of herds were the factors most associated with differences in mean GE-EPDs and genetic trends. Using young sires from other herds or lineages is an effective strategy in animal breeding. By enhancing genetic variability, this approach does not only improve the genetic quality of herds but also accelerates genetic progress in desired traits over time. Therefore, to ensure the success of this strategy, it is crucial that seedstock herds undergo a thorough selection process aimed at maximizing the genetic potential of future generations of beef cattle.

Has genomic selection fulfilled its promise?

Abstract Genomic selection has been increasingly implemented since its initial description in 2001. The initial focus of activity was in the dairy industry in the developed world but it has now been adopted in all major livestock species, and adoption is underway in most agronomically important plants. The promise of genomic selection includes increased accuracy of estimation of genetic merit, earlier estimation of genetic merit, the capacity to make improvements in traits previously less tractable to genetic change, and the capacity to estimate genetic merit in animals without phenotypic records. Each of these benefits has been realised in practice, with the significance of the opportunity varying with species. At the same time, ease of adoption and realisation of benefits has varied, based on three broad parameters: the availability of suitable (and sufficiently cheap) genotyping tools, the prior extent of systematic recording of phenotypes, and the use of estimates of genetic merit in selection. The effectiveness of genomic selection depends on the existence and composition of a genomic reference population, and this is the most significant challenge for any industry, breed or breeding program seeking to achieve the available benefits. The challenge is two-fold: first, to ensure that sufficient of the right phenotypes are recorded on the right animals, and second, flowing directly from the first, is to fund the phenotyping effort – and challenges aspects are ongoing. “Right” in the context of both phenotypes and animals means most informative, or relevant to the breeding objective or goal. No standard approach has been outlined for addressing these challenges, but the review presents some simple conceptual thinking that can assist. Genomic selection also heightens the importance of ensuring maintenance of genetic diversity, and of monitoring changes in both selected and non-selected traits. Genomic selection has fulfilled its promise where applied effectively and is increasing scope for genetic improvement both in rate of progress and trait coverage. At the same time, its dependencies must be addressed in order to the potential benefits to be realised, and this will likely require new organisational and funding arrangements in many situations.

Use of birth weight as a predictor of genetic merit of subsequent growth traits in Duroc and Landrace pigs

Analysis of correlation between the early testable phenotypes of piglets and the final performance of pigs can serve the early selection for breeding. The objectives of this study were to estimate the genetic parameters for birth weight (BtW), age (AGE) and backfat thickness (BF) up to 115 kg BW and to analyse the relationships among these three traits, and to estimate the accuracy of using BtW to predict estimated breeding values (EBVs) of AGE and BF in Landrace and Duroc pigs. Data on 26 614 Landrace and 19 984 Duroc pigs, born between 2001 and 2018, were collected from the core breeding group of a farm. All pigs were recorded for phenotypes including BtW, AGE and BF. The factors affecting these three traits were analysed using R v4.2.0 Software. The population genetic parameters and breeding values of three traits were estimated by using a multitrait animal model based on AI plate of DMU software. Heritabilities for BtW, AGE and BF were moderate to high for Landrace (0.437, 0.282and 0.137, respectively) and Duroc breeds (0.369, 0.279 and 0.148). BtW was genetically correlated with AGE and BF in Landrace (−0.213, 0.037) and Duroc (−0.214, 0.025). AGE was negatively genetically correlated with BF in both Landrace (−0.036) and Duroc (−0.057) pigs. The heritability of BtW, AGE and BF of Landrace pigs and Duroc pigs were 0.148, 0.182 and 0.075 and 0.168, 0.159 and 0.120, respectively, by taking into account of the litter effect. BtW was genetically correlated with AGE and BF in Landrace (−0.094, 0.002) and Duroc (−0.199, −0.052). AGE was negatively genetically correlated with BF in both Landrace (−0.034) and Duroc (−0.153) pigs. The variances between total individual BtW and AGE and BF were then used to predict the EBV of AGE and BF for individuals with AGE or BF phenotypes missing under 10-fold cross−validation. Prediction accuracy was calculated as the Kendall tau-b correlation coefficient between EBVs and EBVs via 10-fold cross-validation. Prediction accuracy for AGE and BF was 0.655 and 0.611 in Landrace, 0.665 and 0.617 in Duroc. After incorporation of the litter effect, the prediction accuracy for AGE and BF increased to 0.690 and 0.665 in Landrace and to 0.705 and 0.649 in Duroc. So, the EBV of AGE and BF phenotypes missing individuals could be predicted by using the available phenotypic data and the easily measured BtW, and litter effect could boost the accuracy of prediction.

Genomic selection optimization in blueberry: Data-driven methods for marker and training population design.

Genomic prediction is a modern approach that uses genome-wide markers to predict the genetic merit of unphenotyped individuals. With the potential to reduce the breeding cycles and increase the selection accuracy, this tool has been designed to rank genotypes and maximize genetic gains. Despite this importance, its practical implementation in breeding programs requires critical allocation of resources for its application in a predictive framework. In this study, we integrated genetic and data-driven methods to allocate resources for phenotyping and genotyping tailored to genomic prediction. To this end, we used a historical blueberry (Vaccinium corymbosun L.) breeding dataset containing more than 3000 individuals, genotyped using probe-based target sequencing and phenotyped for three fruit quality traits over several years. Our contribution in this study is threefold: (i) for the genotyping resource allocation, the use of genetic data-driven methods to select an optimal set of markers slightly improved prediction results for all the traits; (ii) for the long-term implication, we carried out a simulation study and emphasized that data-driven method results in a slight improvement in genetic gain over 30 cycles than random marker sampling; and (iii) for the phenotyping resource allocation, we compared different optimization algorithms to select training population, showing that it can be leveraged to increase predictive performances. Altogether, we provided a data-oriented decision-making approach for breeders by demonstrating that critical breeding decisions associated with resource allocation for genomic prediction can be tackled through a combination of statistics and genetic methods.

Effect of germplasm exchange strategies on genetic gain, homozygosity, and genetic diversity in dairy stud populations: A simulation study

While genomic selection has led to considerable improvements in genetic gain, it has also seemingly led to increased rates of inbreeding and homozygosity, which can negatively affect genetic diversity and the long-term sustainability of dairy populations. Using genotypes from US Holstein animals from 3 distinct stud populations, we performed a simulation study consisting of 10 rounds of selection, with each breeding population composed of 200 males and 2000 females. The investigated selection strategies consisted of selection using true breeding values (TBV), estimated breeding values (EBV), estimated breeding values penalized for the average future genomic inbreeding of progeny (PEN-EBV), or random selection (RAND). We also simulated several germplasm exchange strategies where the germplasm of males from other populations was used for breeding. These strategies included exchanging males based on EBV, PEN-EBV, low genomic future inbreeding of progeny (GFI), or randomly (RAND). Variations of several parameters, such as the correlation between the selection objectives of populations and the size of the exchange, were simulated. Penalizing genetic merit to minimize genomic inbreeding of progeny provided similar genetic gain and reduced the average homozygosity of populations compared with the EBV strategy. Germplasm exchange was found to generally provide long-term benefits to all stud populations. In both the short and long-term, germplasm exchange using the EBV or PEN-EBV strategies provided more cumulative genetic progress than the no-exchange strategy; the amount of long-term genetic progress achieved with germplasm exchange using these strategies was higher for scenarios with a higher genetic correlation between the traits selected by the studs and for a larger size of the exchange. Both the PEN-EBV and GFI exchange strategies allowed decreases in homozygosity and provided significant benefits to genetic diversity compared with other strategies, including larger average minor allele frequencies and smaller proportions of markers near fixation. Overall, this study showed the value of breeding strategies to balance genetic progress and genetic diversity and the benefits of cooperation between studs to ensure the sustainability of their respective breeding programs.

A retrospective analysis of conception per embryo transfer in dairy cattle in South Korea

The bovine embryo production industry has seen significant growth over the past two decades, particularly in the production of in vitro produced embryos. This growth, driven by advancements in cryopreservation, in vitro culture mediums, ovum pick-up (OPU) procedures, ultrasonography devices, and embryo transfer (ET) has been notable. Particularly, ET is crucial for disseminating high genetic merit and amplifying foreign breeds by importing frozen embryos. This retrospective study aimed to assess factors affecting conception per embryo transfer (CPET) in Holstein-Friesian cattle in South Korea from October 2008 to July 2022. We evaluated type of embryo breed, type of embryo production (fresh and frozen; in vitro and in vivo production), recipient conditions including estrus type, corpus luteum quality, parity (nulliparous heifers, primiparous, and multiparous cows), and the daily mean temperature-humidity index (THI) as an index for heat stress. Type of embryo breed and estrus had no significant impact on CPET. However, we observed higher CPET in recipients with good quality corpus luteum, nulliparous heifers, and surrogates receiving fresh in vitro and frozen in vivo embryos. Importantly, CPET was not adversely affected by mild heat stress conditions (up to daily mean THI 76), indicating that using frozen in vivo embryos produced by multiple ovulation embryo transfer and fresh in vitro embryos by OPU-ET can help alleviate the subfertility issues in dairy cattle caused by global warming in Korea.

Estrous activity in nulliparous heifers with divergent genetic merit for fertility traits

This observational study determined the effect of genetic merit for fertility traits on estrous activity and duration and inter-estrous interval in nulliparous dairy heifers. We also compared estrous activity between estrous events with or without conception and determined the effect of genetic merit for fertility traits on age and body weight at time of first detected estrous activity event. Activity monitoring devices (Heatime, SCR Engineers) were fitted to 7-9-mo aged Holstein-Friesian heifers with positive or negative genetic merit for fertility traits (POS FertBV: average +5%, n = 275; NEG FertBV: average -5%, n = 249) and activity data was collected to the end of the first breeding period (15-17-mo). An estrous event was defined as when the activity change index exceeded 19.2 activity units (AU) for ≥6 consecutive h. In total, 2,434 estrous events were identified (POS FertBV: n = 1454; NEG FertBV: n = 980). Estrous event duration was defined as the period when the threshold was first exceeded and when activity dropped below threshold, without another event starting within 24 h of the end of the previous event. This definition included occurrences where activity crossed the threshold multiple times in a day; these were classified as a single estrous event. A second measure, high activity duration, was defined as the total time activity exceeded the threshold. To characterize estrous activity, baseline activity was determined from the previous 7 d of activity. Peak activity and total activity (area under the curve of activity above baseline) were calculated. A fifth variable, inter-estrous interval, was calculated as a proxy for estrous cycle length. Total activity and peak activity were greater in POS than NEG FertBV heifers (total activity: 548 vs 464 AU, SED = 19.6 AU; peak activity: 72 vs 65 AU, standard error of the difference (SED) = 1.5 AU). The POS FertBV group had a mean estrous event duration and high activity duration of 15.1 h and 15.0 h, compared with 14.1 and 14.1 h for the NEG FertBV group (SED = 0.30 and 0.29 h, respectively). Inter-estrous interval did not differ between POS and NEG FertBV heifers (19.5 vs 20.0 d, SED = 0.49). Estrous events associated with conception were shorter than those not associated with conception (mean ± standard error of the mean, high activity duration: 13.0 ± 0.25 h vs 13.9 ± 0.31 h, estrous event duration: 13.1 ± 0.25 h vs 14 ± 0.32 h) and had less total activity (408 ± 15.2 vs 487 ± 18.2 AU). The POS FertBV heifers were more likely to have their first estrous activity event than NEG FertBV heifers by a given age (HR = 1.26, 95% CI = 1.0 to 1.6) or body weight (HR = 1.35, 95% CI = 1.1 to 1.6). The outcomes of this study provide evidence that positive genetic merit for fertility traits is associated with more overt estrous expression. Therefore, estrous expression traits may have potential to be used as earlier-in-life predictors of genetic merit for fertility.

Optimising desired gain indices to maximise selection response.

In plant breeding, we often aim to improve multiple traits at once. However, without knowing the economic value of each trait, it is hard to decide which traits to focus on. This is where "desired gain selection indices" come in handy, which can yield optimal gains in each trait based on the breeder's prioritisation of desired improvements when economic weights are not available. However, they lack the ability to maximise the selection response and determine the correlation between the index and net genetic merit. Here, we report the development of an iterative desired gain selection index method that optimises the sampling of the desired gain values to achieve a targeted or a user-specified selection response for multiple traits. This targeted selection response can be constrained or unconstrained for either a subset or all the studied traits. We tested the method using genomic estimated breeding values (GEBVs) for seven traits in a bread wheat (Triticum aestivum) reference breeding population comprising 3,331 lines and achieved prediction accuracies ranging between 0.29 and 0.47 across the seven traits. The indices were validated using 3,005 double haploid lines that were derived from crosses between parents selected from the reference population. We tested three user-specified response scenarios: a constrained equal weight (INDEX1), a constrained yield dominant weight (INDEX2), and an unconstrained weight (INDEX3). Our method achieved an equivalent response to the user-specified selection response when constraining a set of traits, and this response was much better than the response of the traditional desired gain selection indices method without iteration. Interestingly, when using unconstrained weight, our iterative method maximised the selection response and shifted the average GEBVs of the selection candidates towards the desired direction. Our results show that the method is an optimal choice not only when economic weights are unavailable, but also when constraining the selection response is an unfavourable option.

Selection for feed efficiency improves production traits and digestibility and its relationship to the fecal microbiota in both Large White dam and sire lines

This study evaluated production traits, apparent total tract digestibility (ATTD), and fecal microbiota composition in finishing boars ( n = 207) selected for feed efficiency (FE; low = LFE; high = HFE) based on estimated breeding value for feed conversion ratio (FCR) within a Large White dam and sire genetic lines. Also, the association between gut microbiota with production and digestibility traits was investigated. Regardless of the genetic line, HFE pigs presented low FCR ( P &lt; 0.05), had thinner back fat ( P &lt; 0.05) and had greater loin depth ( P &lt; 0.05) than LFE pigs. Also, HFE pigs had a significantly higher ATTD for Ca and a tendency for a higher crude protein ( P = 0.06) and phosphorous ( P = 0.10) ATTD than LFE pigs. No significant differences in alpha- and beta-diversity measurements for fecal microbiota were seen between FE groups in each genetic line. The genera Lactobacillus and Prevotella were associated ( P &lt; 0.01) with some growth performance (e.g., feed intake), carcass traits (e.g., backfat thickness), and nutrient digestibility (e.g., Ca). In conclusion, HFE pigs had favourable production traits and higher digestibility of key nutrients than LFE pigs; however, the bacterial genera were associated with phenotypic traits but not by genetic merit.

A comparison of genomic and phenomic selection methods for yield prediction in Coffea canephora

AbstractGenomic prediction has been proposed as the standard method to predict the genetic merit of unphenotyped individuals. Despite the promising results reported in the plant breeding literature, its routine implementation remains difficult for some crops. This is the case with Coffea canephora, in which costs and availability of molecular tools are major challenges for most breeding programs. To circumvent this, the use of near‐infrared spectroscopy (NIR) has been recently proposed as an alternative to complement marker‐assisted selection. The so‐called phenomic selection relies on the reflectance spectrum to capture similarities between individuals and emerges as a valid approach for prediction. With promising results reported in multiple annual crops, we hypothesize that phenomic prediction could be a cost‐efficient approach to incorporate into a practical coffee breeding program. To test it, we relied on a diverse population of C. canephora, evaluated for yield production, in two geographical locations over four harvest seasons. Our contributions in this paper are twofold: (i) We compared phenomic and genomic selection results, and showed large predictive abilities when NIR is used as a predictor for within and across‐location predictions, and (ii) we presented a critical view of how both information sets could be combined into a contemporaneous coffee breeding program. Altogether, our results show how multi‐omic information could be integrated in the same framework to leverage genetic gains in the long term.