Dataset Information

A comparison of five methods to predict genomic breeding values of dairy bulls from genome-wide SNP markers.

ABSTRACT: BACKGROUND: Genomic selection (GS) uses molecular breeding values (MBV) derived from dense markers across the entire genome for selection of young animals. The accuracy of MBV prediction is important for a successful application of GS. Recently, several methods have been proposed to estimate MBV. Initial simulation studies have shown that these methods can accurately predict MBV. In this study we compared the accuracies and possible bias of five different regression methods in an empirical application in dairy cattle. METHODS: Genotypes of 7,372 SNP and highly accurate EBV of 1,945 dairy bulls were used to predict MBV for protein percentage (PPT) and a profit index (Australian Selection Index, ASI). Marker effects were estimated by least squares regression (FR-LS), Bayesian regression (Bayes-R), random regression best linear unbiased prediction (RR-BLUP), partial least squares regression (PLSR) and nonparametric support vector regression (SVR) in a training set of 1,239 bulls. Accuracy and bias of MBV prediction were calculated from cross-validation of the training set and tested against a test team of 706 young bulls. RESULTS: For both traits, FR-LS using a subset of SNP was significantly less accurate than all other methods which used all SNP. Accuracies obtained by Bayes-R, RR-BLUP, PLSR and SVR were very similar for ASI (0.39-0.45) and for PPT (0.55-0.61). Overall, SVR gave the highest accuracy.All methods resulted in biased MBV predictions for ASI, for PPT only RR-BLUP and SVR predictions were unbiased. A significant decrease in accuracy of prediction of ASI was seen in young test cohorts of bulls compared to the accuracy derived from cross-validation of the training set. This reduction was not apparent for PPT. Combining MBV predictions with pedigree based predictions gave 1.05 - 1.34 times higher accuracies compared to predictions based on pedigree alone. Some methods have largely different computational requirements, with PLSR and RR-BLUP requiring the least computing time. CONCLUSIONS: The four methods which use information from all SNP namely RR-BLUP, Bayes-R, PLSR and SVR generate similar accuracies of MBV prediction for genomic selection, and their use in the selection of immediate future generations in dairy cattle will be comparable. The use of FR-LS in genomic selection is not recommended.

SUBMITTER: Moser G

PROVIDER: S-EPMC2814805 | biostudies-other | 2009

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

A comparison of five methods to predict genomic breeding values of dairy bulls from genome-wide SNP markers.

Moser Gerhard G Tier Bruce B Crump Ron E RE Khatkar Mehar S MS Raadsma Herman W HW

Genetics, selection, evolution : GSE 20091231

<h4>Background</h4>Genomic selection (GS) uses molecular breeding values (MBV) derived from dense markers across the entire genome for selection of young animals. The accuracy of MBV prediction is important for a successful application of GS. Recently, several methods have been proposed to estimate MBV. Initial simulation studies have shown that these methods can accurately predict MBV. In this study we compared the accuracies and possible bias of five different regression methods in an empirica ...[more]

PMID: 20043835

Similar Datasets

Project description:BackgroundAt the current price, the use of high-density single nucleotide polymorphisms (SNP) genotyping assays in genomic selection of dairy cattle is limited to applications involving elite sires and dams. The objective of this study was to evaluate the use of low-density assays to predict direct genomic value (DGV) on five milk production traits, an overall conformation trait, a survival index, and two profit index traits (APR, ASI).MethodsDense SNP genotypes were available for 42,576 SNP for 2,114 Holstein bulls and 510 cows. A subset of 1,847 bulls born between 1955 and 2004 was used as a training set to fit models with various sets of pre-selected SNP. A group of 297 bulls born between 2001 and 2004 and all cows born between 1992 and 2004 were used to evaluate the accuracy of DGV prediction. Ridge regression (RR) and partial least squares regression (PLSR) were used to derive prediction equations and to rank SNP based on the absolute value of the regression coefficients. Four alternative strategies were applied to select subset of SNP, namely: subsets of the highest ranked SNP for each individual trait, or a single subset of evenly spaced SNP, where SNP were selected based on their rank for ASI, APR or minor allele frequency within intervals of approximately equal length.ResultsRR and PLSR performed very similarly to predict DGV, with PLSR performing better for low-density assays and RR for higher-density SNP sets. When using all SNP, DGV predictions for production traits, which have a higher heritability, were more accurate (0.52-0.64) than for survival (0.19-0.20), which has a low heritability. The gain in accuracy using subsets that included the highest ranked SNP for each trait was marginal (5-6%) over a common set of evenly spaced SNP when at least 3,000 SNP were used. Subsets containing 3,000 SNP provided more than 90% of the accuracy that could be achieved with a high-density assay for cows, and 80% of the high-density assay for young bulls.ConclusionsAccurate genomic evaluation of the broader bull and cow population can be achieved with a single genotyping assays containing ~ 3,000 to 5,000 evenly spaced SNP.

Project description:BackgroundFor infectious diseases, the probability that an animal gets infected depends on its own susceptibility, and on the number of infectious herd mates and their infectivity. Together with the duration of the infectious period, susceptibility and infectivity determine the basic reproduction ratio of the disease ([Formula: see text]). [Formula: see text] is the average number of secondary cases caused by a typical infectious individual in an otherwise uninfected population. An infectious disease dies out when [Formula: see text] is lower than 1. Thus, breeding strategies that aim at reducing disease prevalence should focus on reducing [Formula: see text], preferably to a value lower than 1. In animal breeding, however, [Formula: see text] has received little attention. Here, we estimate the additive genetic variance in host susceptibility, host infectivity, and [Formula: see text] for the endemic claw disease digital dermatitis (DD) in Holstein Friesian dairy cattle, and estimate genomic breeding values (GEBV) for these traits. We recorded DD disease status of both hind claws of 1513 cows from 12 Dutch dairy farms, every 2 weeks, 11 times. The genotype data consisted of 75,904 single nucleotide polymorphisms (SNPs) for 1401 of the cows. We modelled the probability that a cow got infected between recordings, and compared four generalized linear mixed models. All models included a genetic effect for susceptibility; Models 2 and 4 also included a genetic effect for infectivity, while Models 1 and 2 included a farm*period interaction. We corrected for variation in exposure to infectious herd mates via an offset.ResultsGEBV for [Formula: see text] from the model that included genetic effects for susceptibility only had an accuracy of ~ 0.39 based on cross-validation between farms, which is very high given the limited amount of data and the complexity of the trait. Models with a genetic effect for infectivity showed a larger bias, but also a slightly higher accuracy of GEBV. Additive genetic standard deviation for [Formula: see text] was large, i.e. ~ 1.17, while the mean [Formula: see text] was 2.36.ConclusionsGEBV for [Formula: see text] showed substantial variation. The mean [Formula: see text] was only about one genetic standard deviation greater than 1. These results suggest that lowering DD prevalence by selective breeding is promising.

Project description:Genetic selection for higher productivity increased dairy sheep susceptibility to diseases and environmental stressors, challenging their health and welfare status and production efficiency. Improving resilience to such stressors can enhance their ability to face these challenges without compromising productivity. Our objective was to estimate genomic heritability and perform genome-wide association studies (GWAS) to detect SNPs and candidate genes associated with three proxy traits for resilience (milk somatic cell count-SCC, lactation persistency-LP, body condition score-BCS) of Chios and Frizarta dairy ewes. We used genome-wide genotypes of 317 Chios and 346 Frizarta ewes. Individual records of milk yield and BCS, and milk samples were collected monthly for two consecutive milking periods; samples were analyzed to determine SCC. The LP was calculated as the regression coefficient of daily milk yield on days from lambing. Within breed, variance components analyses and GWAS were performed using genomic relatedness matrices in single-trait animal linear mixed models. Genomic-based heritability estimates were relatively high (BCS: h2 = 0.54 and 0.55, SCC: h2 = 0.25 and 0.38, LP: h2 = 0.43 and 0.45, for Chios and Frizarta ewes, respectively), compared to previous pedigree-based studies. The GWAS revealed 7 novel SNPs associated with the studied traits; one genome-wide and two suggestive significant SNPs for SCC (Frizarta: rs403061409, rs424064526 and rs428540973, on chromosomes 9, 1 and 12, respectively), one suggestive significant SNP for BCS (Chios: rs424834097 on chromosome 4) and three suggestive significant SNPs for LP (Frizarta: rs193632931 and rs412648955 on chromosomes 1 and 6, Chios: rs428128299 on chromosome 3). Nineteen candidate genes were detected: two for BCS (Chios: POT1, TMEM229A), thirteen for SCC (Frizarta: NTAQ1, ZHX1, ZHX2, LOC101109545, HAS2, DERL1, FAM83A, ATAD2, RBP7, FSTL1, CD80, HCLS1, GSK3B) and four for LP (Frizarta: GRID2, FAIM, CEP70-Chios: GRIP1). Present results show that resilience in the studied dairy sheep breeds is heritable and advance existing knowledge on the genomic background of SCC, LP, and BCS. Future research will quantify effects of different alleles of significant SNPs on the studied traits and search for possible correlations among traits to facilitate their effective incorporation in breeding programs aiming to improve resilience.

Project description:In crossbreeding programs, various genomic prediction models have been proposed for using phenotypic records of crossbred animals to increase the selection response for crossbred performance in purebred animals. A possible model is a model that assumes identical single nucleotide polymorphism (SNP) effects for the crossbred performance trait across breeds (ASGM). Another model is a genomic model that assumes breed-specific effects of SNP alleles (BSAM) for crossbred performance. The aim of this study was to derive and validate equations for predicting the reliability of estimated genomic breeding values for crossbred performance in both these models. Prediction equations were derived for situations when all (phenotyping and) genotyping data have already been collected, i.e. based on the genetic evaluation model, and for situations when all genotyping data are not yet available, i.e. when designing breeding programs.When all genotyping data are available, prediction equations are based on selection index theory. Without availability of all genotyping data, prediction equations are based on population parameters (e.g., heritability of the traits involved, genetic correlation between purebred and crossbred performance, effective number of chromosome segments). Validation of the equations for predicting the reliability of genomic breeding values without all genotyping data was performed based on simulated data of a two-way crossbreeding program, using either two closely-related breeds, or two unrelated breeds, to produce crossbred animals. The proposed equations can be used for an easy comparison of the reliability of genomic estimated breeding values across many scenarios, especially if all genotyping data are available. We show that BSAM outperforms ASGM for a specific breed, if the effective number of chromosome segments that originate from this breed and are shared by selection candidates of this breed and crossbred reference animals is less than half the effective number of all chromosome segments that are independently segregating in the same animals.The derived equations can be used to predict the reliability of genomic estimated breeding values for crossbred performance using ASGM or BSAM in many scenarios, and are thus useful to optimize the design of breeding programs. Scenarios can vary in terms of the genetic correlation between purebred and crossbred performances, heritabilities, number of reference animals, or distance between breeds.

Project description:Genomic selection has been widely implemented in dairy cattle breeding when the aim is to improve performance of purebred animals. In pigs, however, the final product is a crossbred animal. This may affect the efficiency of methods that are currently implemented for dairy cattle. Therefore, the objective of this study was to determine the accuracy of predicted breeding values in crossbred pigs using purebred genomic and phenotypic data. A second objective was to compare the predictive ability of SNPs when training is done in either single or multiple populations for four traits: age at first insemination (AFI); total number of piglets born (TNB); litter birth weight (LBW); and litter variation (LVR). We performed marker-based and pedigree-based predictions. Within-population predictions for the four traits ranged from 0.21 to 0.72. Multi-population prediction yielded accuracies ranging from 0.18 to 0.67. Predictions across purebred populations as well as predicting genetic merit of crossbreds from their purebred parental lines for AFI performed poorly (not significantly different from zero). In contrast, accuracies of across-population predictions and accuracies of purebred to crossbred predictions for LBW and LVR ranged from 0.08 to 0.31 and 0.11 to 0.31, respectively. Accuracy for TNB was zero for across-population prediction, whereas for purebred to crossbred prediction it ranged from 0.08 to 0.22. In general, marker-based outperformed pedigree-based prediction across populations and traits. However, in some cases pedigree-based prediction performed similarly or outperformed marker-based prediction. There was predictive ability when purebred populations were used to predict crossbred genetic merit using an additive model in the populations studied. AFI was the only exception, indicating that predictive ability depends largely on the genetic correlation between PB and CB performance, which was 0.31 for AFI. Multi-population prediction was no better than within-population prediction for the purebred validation set. Accuracy of prediction was very trait-dependent.

Project description:BackgroundOur aim was to identify genomic regions via genome-wide association studies (GWAS) to improve the predictability of genetic merit in Holsteins for 10 calving and 28 body conformation traits. Animals were genotyped using the Illumina Bovine 50 K BeadChip and imputed to the Illumina BovineHD BeadChip (HD). GWAS were performed on 601,717 real and imputed single nucleotide polymorphism (SNP) genotypes using a single-SNP mixed linear model on 4841 Holstein bulls with breeding value predictions and followed by gene identification and in silico functional analyses. The association results were further validated using five scenarios with different numbers of SNPs.ResultsSeven hundred and eighty-two SNPs were significantly associated with calving performance at a genome-wise false discovery rate (FDR) of 5%. Most of these significant SNPs were on chromosomes 18 (71.9%), 17 (7.4%), 5 (6.8%) and 7 (2.4%) and mapped to 675 genes, among which 142 included at least one significant SNP and 532 were nearby one (100 kbp). For body conformation traits, 607 SNPs were significant at a genome-wise FDR of 5% and most of them were located on chromosomes 5 (30%), 18 (27%), 20 (13%), 6 (6%), 7 (5%), 14 (5%) and 13 (3%). SNP enrichment functional analyses for calving traits at a FDR of 1% suggested potential biological processes including musculoskeletal movement, meiotic cell cycle, oocyte maturation and skeletal muscle contraction. Furthermore, pathway analyses suggested potential pathways associated with calving performance traits including tight junction, oxytocin signaling, and MAPK signaling (P < 0.10). The prediction ability of the 1206 significant SNPs was between 78 and 83% of the prediction ability of the BovineSNP50 SNPs for calving performance traits and between 35 and 79% for body conformation traits.ConclusionsVarious SNPs that are significantly associated with calving performance are located within or nearby genes with potential roles in tight junction, oxytocin signaling, and MAPK signaling. Combining the significant SNPs or SNPs within or nearby gene(s) from the HD panel with the BovineSNP50 panel yielded a marginal increase in the accuracy of prediction of genomic estimated breeding values for all traits compared to the use of the BovineSNP50 panel alone.

Dataset Information

A comparison of five methods to predict genomic breeding values of dairy bulls from genome-wide SNP markers.

Publications

A comparison of five methods to predict genomic breeding values of dairy bulls from genome-wide SNP markers.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets