Genome-wide association study on Fourier transform infrared milk spectra for two Danish dairy cattle breeds.

BACKGROUND: Infrared spectral analysis of milk is cheap, fast, and accurate. Infrared light interacts with chemical bonds present inside the milk, which means that Fourier transform infrared milk spectra are a reflection of the chemical composition of milk. Heritability of Fourier transform infrared milk spectra has been analysed previously. Further genetic analysis of Fourier transform infrared milk spectra could give us a better insight in the genes underlying milk composition. Breed influences milk composition, yet not much is known about the effect of breed on Fourier transform infrared milk spectra. Improved understanding of the effect of breed on Fourier transform infrared milk spectra could enhance efficient application of Fourier transform infrared milk spectra. The aim of this study is to perform a genome wide association study on a selection of wavenumbers for Danish Holstein and Danish Jersey. This will improve our understanding of the genetics underlying milk composition in these two dairy cattle breeds. RESULTS: For each breed separately, fifteen wavenumbers were analysed. Overall, more quantitative trait loci were observed for Danish Jersey compared to Danish Holstein. For both breeds, the majority of the wavenumbers was most strongly associated to a genomic region on BTA 14 harbouring DGAT1. Furthermore, for both breeds most quantitative trait loci were observed for wavenumbers that interact with the chemical bond C-O. For Danish Jersey, wavenumbers that interact with C-H were associated to genes that are involved in fatty acid synthesis, such as AGPAT3, AGPAT6, PPARGC1A, SREBF1, and FADS1. For wavenumbers which interact with -OH, associations were observed to genomic regions that have been linked to alpha-lactalbumin. CONCLUSIONS: The current study identified many quantitative trait loci that underlie Fourier transform infrared milk spectra, and thus milk composition. Differences were observed between groups of wavenumbers that interact with different chemical bonds. Both overlapping and different QTL were observed for Danish Holstein and Danish Jersey.

Genetic analysis of Fourier transform infrared milk spectra in Danish Holstein and Danish Jersey.

Fourier transform infrared milk spectral data are routinely used for milk quality control and have been revealed to be driven by genetics. This study aimed to (1) estimate heritability for 1,060 wavenumbers in the infrared region from 5,008 to 925 cm-1, (2) estimate genomic correlations between wavenumbers with increased heritability, and (3) compare results between Danish Holstein and Danish Jersey cows. For Danish Holstein, 3,275 cows and 19,656 milk records were available. For Danish Jersey, 3,408 cows and 20,228 milk records were available. We used a hierarchical mixed model, with a Bayesian approach. Heritability of individual wavenumbers ranged from 0.00 to 0.31 in Danish Holstein, and from 0.00 to 0.30 in Danish Jersey. Genomic correlation was calculated between 15 selected wavenumbers, and varied from weak to very strong, in both Danish Holstein and Danish Jersey (0.03 to 0.97, and -0.11 to -0.97). Within the 15 selected wavenumbers, a subdivision into 2 groups of wavenumbers was observed, where genomic correlations were negative between groups, and positive within groups. Heritability and genomic correlations were higher in Danish Holstein compared with Danish Jersey, but followed a similar pattern in both breeds. Breed differences were most pronounced in the mid-infrared region that interacts with lactose and the spectral region that interacts with protein. In conclusion, heritability for individual wavenumbers of Fourier transform milk spectra was moderate, and strong genomic correlations were observed between wavenumbers across the spectrum. Heritability and genomic correlations were higher in Danish Holstein, with the strongest breed differences showing in spectral regions interacting with protein or lactose.

Bivariate threshold models for genetic evaluation of susceptibility to and ability to recover from mastitis in Danish Holstein cows.

Mastitis in dairy cows is an unavoidable problem and genetic variation in recovery from mastitis, in addition to susceptibility, is therefore of interest. Genetic parameters for susceptibility to and recovery from mastitis were estimated for Danish Holstein-Friesian cows using data from automatic milking systems equipped with online somatic cell count measuring units. The somatic cell count measurements were converted to elevated mastitis risk, a continuous variable [on a (0-1) scale] indicating the risk of mastitis. Risk values >0.6 were assumed to indicate that a cow had mastitis. For each cow and lactation, the sequence of health states (mastitic or healthy) was converted to a weekly transition: 0 if the cow stayed within the same state and 1 if the cow changed state. The result was 2 series of transitions: one for healthy to diseased (HD, to model mastitis susceptibility) and the other for diseased to healthy (DH, to model recovery ability). The 2 series of transitions were analyzed with bivariate threshold models, including several systematic effects and a function of time. The model included effects of herd, parity, herd-test-week, permanent environment (to account for the repetitive nature of transition records from a cow) plus two time-varying effects (lactation stage and time within episode). In early lactation, there was an increased risk of getting mastitis but the risk remained stable afterwards. Mean recovery rate was 45% per lactation. Heritabilities were 0.07 [posterior mean of standard deviations (PSD) = 0.03] for HD and 0.08 (PSD = 0.03) for DH. The genetic correlation between HD and DH has a posterior mean of -0.83 (PSD = 0.13). Although susceptibility and recovery from mastitis are strongly negatively correlated, recovery can be considered as a new trait for selection.

Decomposing genomic variance using information from GWA, GWE and eQTL analysis.

A commonly used procedure in genome-wide association (GWA), genome-wide expression (GWE) and expression quantitative trait locus (eQTL) analyses is based on a bottom-up experimental approach that attempts to individually associate molecular variants with complex traits. Top-down modeling of the entire set of genomic data and partitioning of the overall variance into subcomponents may provide further insight into the genetic basis of complex traits. To test this approach, we performed a whole-genome variance components analysis and partitioned the genomic variance using information from GWA, GWE and eQTL analyses of growth-related traits in a mouse F2 population. We characterized the mouse trait genetic architecture by ordering single nucleotide polymorphisms (SNPs) based on their P-values and studying the areas under the curve (AUCs). The observed traits were found to have a genomic variance profile that differed significantly from that expected of a trait under an infinitesimal model. This situation was particularly true for both body weight and body fat, for which the AUCs were much higher compared with that of glucose. In addition, SNPs with a high degree of trait-specific regulatory potential (SNPs associated with subset of transcripts that significantly associated with a specific trait) explained a larger proportion of the genomic variance than did SNPs with high overall regulatory potential (SNPs associated with transcripts using traditional eQTL analysis). We introduced AUC measures of genomic variance profiles that can be used to quantify relative importance of SNPs as well as degree of deviation of a trait's inheritance from an infinitesimal model. The shape of the curve aids global understanding of traits: The steeper the left-hand side of the curve, the fewer the number of SNPs controlling most of the phenotypic variance.

Short communication: Multi-trait estimation of genetic parameters for milk protein composition in the Danish Holstein.

Genetic parameters were estimated for the major milk proteins using bivariate and multi-trait models based on genomic relationships between animals. The analyses included, apart from total protein percentage, αS1-casein (CN), αS2-CN, ß-CN, κ-CN, α-lactalbumin, and ß-lactoglobulin, as well as the posttranslational sub-forms of glycosylated κ-CN and αS1-CN-8P (phosphorylated). Standard errors of the estimates were used to compare the models. In total, 650 Danish Holstein cows across 4 parities and days in milk ranging from 9 to 481d were selected from 21 herds. The multi-trait model generally resulted in lower standard errors of heritability estimates, suggesting that genetic parameters can be estimated with high accuracy using multi-trait analyses with genomic relationships for scarcely recorded traits. The heritability estimates from the multi-trait model ranged from low (0.05 for ß-CN) to high (0.78 for κ-CN). Genetic correlations between the milk proteins and the total milk protein percentage were generally low, suggesting the possibility to alter protein composition through selective breeding with little effect on total milk protein percentage.

Estimation of indirect genetic effects in group-housed mink (Neovison vison) should account for systematic interactions either due to kin or sex.

Social interactions among individuals are abundant, both in wild and in domestic populations. With social interactions, the genes of an individual may affect the trait values of other individuals, a phenomenon known as indirect genetic effects (IGEs). IGEs can be estimated using linear mixed models. Most IGE models assume that individuals interact equally to all group mates irrespective of relatedness. Kin selection theory, however, predicts that an individual will interact differently with family members versus non-family members. Here, we investigate kin- and sex-specific non-genetic social interactions in group-housed mink. Furthermore, we investigated whether systematic non-genetic interactions between kin or individuals of the same sex influence the estimates of genetic parameters. As a second objective, we clarify the relationship between estimates of the traditional IGE model and a family-based IGE model proposed in a previous study. Our results indicate that male siblings in mink show different non-genetic interactions than female siblings in mink and that this may impact the estimation of genetic parameters. Moreover, we have shown how estimates from a family-based IGE model can be translated to the ordinary direct-indirect model and vice versa. We find no evidence for genetic differences in interactions among related versus unrelated mink.

Genomic prediction of growth in pigs based on a model including additive and dominance effects.

Independent of whether prediction is based on pedigree or genomic information, the focus of animal breeders has been on additive genetic effects or 'breeding values'. However, when predicting phenotypes rather than breeding values of an animal, models that account for both additive and dominance effects might be more accurate. Our aim with this study was to compare the accuracy of predicting phenotypes using a model that accounts for only additive effects (MA) and a model that accounts for both additive and dominance effects simultaneously (MAD). Lifetime daily gain (DG) was evaluated in three pig populations (1424 Pietrain, 2023 Landrace, and 2157 Large White). Animals were genotyped using the Illumina SNP60K Beadchip and assigned to either a training data set to estimate the genetic parameters and SNP effects, or to a validation data set to assess the prediction accuracy. Models MA and MAD applied random regression on SNP genotypes and were implemented in the program Bayz. The additive heritability of DG across the three populations and the two models was very similar at approximately 0.26. The proportion of phenotypic variance explained by dominance effects ranged from 0.04 (Large White) to 0.11 (Pietrain), indicating that importance of dominance might be breed-specific. Prediction accuracies were higher when predicting phenotypes using total genetic values (sum of breeding values and dominance deviations) from the MAD model compared to using breeding values from both MA and MAD models. The highest increase in accuracy (from 0.195 to 0.222) was observed in the Pietrain, and the lowest in Large White (from 0.354 to 0.359). Predicting phenotypes using total genetic values instead of breeding values in purebred data improved prediction accuracy and reduced the bias of genomic predictions. Additional benefit of the method is expected when applied to predict crossbred phenotypes, where dominance levels are expected to be higher.

Genome-wide association study reveals novel loci for litter size and its variability in a Large White pig population.

BACKGROUND: In many traits, not only individual trait levels are under genetic control, but also the variation around that level. In other words, genotypes do not only differ in mean, but also in (residual) variation around the genotypic mean. New statistical methods facilitate gaining knowledge on the genetic architecture of complex traits such as phenotypic variability. Here we study litter size (total number born) and its variation in a Large White pig population using a Double Hierarchical Generalized Linear model, and perform a genome-wide association study using a Bayesian method. RESULTS: In total, 10 significant single nucleotide polymorphisms (SNPs) were detected for total number born (TNB) and 9 SNPs for variability of TNB (varTNB). Those SNPs explained 0.83 % of genetic variance in TNB and 1.44 % in varTNB. The most significant SNP for TNB was detected on Sus scrofa chromosome (SSC) 11. A possible candidate gene for TNB is ENOX1, which is involved in cell growth and survival. On SSC7, two possible candidate genes for varTNB are located. The first gene is coding a swine heat shock protein 90 (HSPCB = Hsp90), which is a well-studied gene stabilizing morphological traits in Drosophila and Arabidopsis. The second gene is VEGFA, which is activated in angiogenesis and vasculogenesis in the fetus. Furthermore, the genetic correlation between additive genetic effects on TNB and on its variation was 0.49. This indicates that the current selection to increase TNB will also increase the varTNB. CONCLUSIONS: To the best of our knowledge, this is the first study reporting SNPs associated with variation of a trait in pigs. Detected genomic regions associated with varTNB can be used in genomic selection to decrease varTNB, which is highly desirable to avoid very small or very large litters in pigs. However, the percentage of variance explained by those regions was small. The SNPs detected in this study can be used as indication for regions in the Sus scrofa genome involved in maintaining low variability of litter size, but further studies are needed to identify the causative loci.

Quantitative trait loci markers derived from whole genome sequence data increases the reliability of genomic prediction.

This study investigated the effect on the reliability of genomic prediction when a small number of significant variants from single marker analysis based on whole genome sequence data were added to the regular 54k single nucleotide polymorphism (SNP) array data. The extra markers were selected with the aim of augmenting the custom low-density Illumina BovineLD SNP chip (San Diego, CA) used in the Nordic countries. The single-marker analysis was done breed-wise on all 16 index traits included in the breeding goals for Nordic Holstein, Danish Jersey, and Nordic Red cattle plus the total merit index itself. Depending on the trait's economic weight, 15, 10, or 5 quantitative trait loci (QTL) were selected per trait per breed and 3 to 5 markers were selected to tag each QTL. After removing duplicate markers (same marker selected for more than one trait or breed) and filtering for high pairwise linkage disequilibrium and assaying performance on the array, a total of 1,623 QTL markers were selected for inclusion on the custom chip. Genomic prediction analyses were performed for Nordic and French Holstein and Nordic Red animals using either a genomic BLUP or a Bayesian variable selection model. When using the genomic BLUP model including the QTL markers in the analysis, reliability was increased by up to 4 percentage points for production traits in Nordic Holstein animals, up to 3 percentage points for Nordic Reds, and up to 5 percentage points for French Holstein. Smaller gains of up to 1 percentage point was observed for mastitis, but only a 0.5 percentage point increase was seen for fertility. When using a Bayesian model accuracies were generally higher with only 54k data compared with the genomic BLUP approach, but increases in reliability were relatively smaller when QTL markers were included. Results from this study indicate that the reliability of genomic prediction can be increased by including markers significant in genome-wide association studies on whole genome sequence data alongside the 54k SNP set.

Indirect genetic effects and kin recognition: estimating IGEs when interactions differ between kin and strangers.

Social interactions among individuals are widespread, both in natural and domestic populations. As a result, trait values of individuals may be affected by genes in other individuals, a phenomenon known as indirect genetic effects (IGEs). IGEs can be estimated using linear mixed models. The traditional IGE model assumes that an individual interacts equally with all its partners, whether kin or strangers. There is abundant evidence, however, that individuals behave differently towards kin as compared with strangers, which agrees with predictions from kin-selection theory. With a mix of kin and strangers, therefore, IGEs estimated from a traditional model may be incorrect, and selection based on those estimates will be suboptimal. Here we investigate whether genetic parameters for IGEs are statistically identifiable in group-structured populations when IGEs differ between kin and strangers, and develop models to estimate such parameters. First, we extend the definition of total breeding value and total heritable variance to cases where IGEs depend on relatedness. Next, we show that the full set of genetic parameters is not identifiable when IGEs differ between kin and strangers. Subsequently, we present a reduced model that yields estimates of the total heritable effects on kin, on non-kin and on all social partners of an individual, as well as the total heritable variance for response to selection. Finally we discuss the consequences of analysing data in which IGEs depend on relatedness using a traditional IGE model, and investigate group structures that may allow estimation of the full set of genetic parameters when IGEs depend on kin.

Comparison of genomic predictions using genomic relationship matrices built with different weighting factors to account for locus-specific variances.

Various models have been used for genomic prediction. Bayesian variable selection models often predict more accurate genomic breeding values than genomic BLUP (GBLUP), but GBLUP is generally preferred for routine genomic evaluations because of low computational demand. The objective of this study was to achieve the benefits of both models using results from Bayesian models and genome-wide association studies as weights on single nucleotide polymorphism (SNP) markers when constructing the genomic matrix (G-matrix) for genomic prediction. The data comprised 5,221 progeny-tested bulls from the Nordic Holstein population. The animals were genotyped using the Illumina Bovine SNP50 BeadChip (Illumina Inc., San Diego, CA). Weighting factors in this investigation were the posterior SNP variance, the square of the posterior SNP effect, and the corresponding minus base-10 logarithm of the marker association P-value [-log10(P)] of a t-test obtained from the analysis using a Bayesian mixture model with 4 normal distributions, the square of the estimated SNP effect, and the corresponding -log10(P) of a t-test obtained from the analysis using a classical genome-wide association study model (linear regression model). The weights were derived from the analysis based on data sets that were 0, 1, 3, or 5 yr before performing genomic prediction. In building a G-matrix, the weights were assigned either to each marker (single-marker weighting) or to each group of approximately 5 to 150 markers (group-marker weighting). The analysis was carried out for milk yield, fat yield, protein yield, fertility, and mastitis. Deregressed proofs (DRP) were used as response variables to predict genomic estimated breeding values (GEBV). Averaging over the 5 traits, the Bayesian model led to 2.0% higher reliability of GEBV than the GBLUP model with an original unweighted G-matrix. The superiority of using a GBLUP with weighted G-matrix over GBLUP with an original unweighted G-matrix was the largest when using a weighting factor of posterior variance, resulting in 1.7 percentage points higher reliability. The second best weighting factors were -log10 (P-value) of a t-test corresponding to the square of the posterior SNP effect from the Bayesian model and -log10 (P-value) of a t-test corresponding to the square of the estimated SNP effect from the linear regression model, followed by the square of estimated SNP effect and the square of the posterior SNP effect. In addition, group-marker weighting performed better than single-marker weighting in terms of reducing bias of GEBV, and also slightly increased prediction reliability. The differences between weighting factors and scenarios were larger in prediction bias than in prediction accuracy. Finally, weights derived from a data set having a lag up to 3 yr did not reduce reliability of GEBV. The results indicate that posterior SNP variance estimated from a Bayesian mixture model is a good alternative weighting factor, and common weights on group markers with a size of 30 markers is a good strategy when using markers of the 50,000-marker (50K) chip. In a population with gradually increasing reference data, the weights can be updated once every 3 yr.

Model comparison on genomic predictions using high-density markers for different groups of bulls in the Nordic Holstein population.

This study compared genomic predictions based on imputed high-density markers (~777,000) in the Nordic Holstein population using a genomic BLUP (GBLUP) model, 4 Bayesian exponential power models with different shape parameters (0.3, 0.5, 0.8, and 1.0) for the exponential power distribution, and a Bayesian mixture model (a mixture of 4 normal distributions). Direct genomic values (DGV) were estimated for milk yield, fat yield, protein yield, fertility, and mastitis, using deregressed proofs (DRP) as response variable. The validation animals were split into 4 groups according to their genetic relationship with the training population. Groupsmgs had both the sire and the maternal grandsire (MGS), Groupsire only had the sire, Groupmgs only had the MGS, and Groupnon had neither the sire nor the MGS in the training population. Reliability of DGV was measured as the squared correlation between DGV and DRP divided by the reliability of DRP for the bulls in validation data set. Unbiasedness of DGV was measured as the regression of DRP on DGV. The results indicated that DGV were more accurate and less biased for animals that were more related to the training population. In general, the Bayesian mixture model and the exponential power model with shape parameter of 0.30 led to higher reliability of DGV than did the other models. The differences between reliabilities of DGV from the Bayesian models and the GBLUP model were statistically significant for some traits. We observed a tendency that the superiority of the Bayesian models over the GBLUP model was more profound for the groups having weaker relationships with training population. Averaged over the 5 traits, the Bayesian mixture model improved the reliability of DGV by 2.0 percentage points for Groupsmgs, 2.7 percentage points for Groupsire, 3.3 percentage points for Groupmgs, and 4.3 percentage points for Groupnon compared with GBLUP. The results showed that a Bayesian model with intense shrinkage of the explanatory variable, such as the Bayesian mixture model and the Bayesian exponential power model with shape parameter of 0.30, can improve genomic predictions using high-density markers.

Genotype by environment interaction for somatic cell score across bulk milk somatic cell count and days in milk.

The objective of this paper was to investigate the importance of a genotype x environment interaction (G x E) for somatic cell score (SCS) across levels of bulk milk somatic cell count (BMSCC), number of days in milk (DIM), and their interaction. Variance components were estimated with a model including random regressions for each sire on herd test-day BMSCC, DIM, and the interaction of BMSCC and DIM. The analyzed data set contained 344,029 test-day records of 24,125 cows, sired by 182 bulls, in 461 herds comprising 13,563 herd test-days. In early lactation, considerable G x E effects were detected for SCS, indicated by 3-fold higher genetic variance for SCS at high BMSCC compared with SCS at low BMSCC, and a genetic correlation of 0.72 between SCS at low and at high BMSCC. Estimated G x E effects were smaller during late lactation. Genetic correlations between SCS at the same level of BMSCC, across DIM, were between 0.43 and 0.89. The lowest genetic correlation between SCS measures on any 2 possible combinations of BMSCC and DIM was 0.42. Correlated responses in SCS across BMSCC and DIM were, on some occasions, less than half the direct response to selection in the response environment. Responses to selection were reasonably high among environments in the second half of the lactation, whereas responses to selection between environments early and late in lactation tended to be low. Selection for reduced SCS yielded the highest direct response early in lactation at high BMSCC.

A note on the estimation of the effective number of additive and dominant loci contributing to quantitative variation.

A method of estimating the number of loci contributing to quantitative variation has been proposed by S. Wright in 1921. The method makes use of the means of inbred lines and the variances of their F1, F2 and backcrosses. The method has been extended to crosses between outbreeding populations by R. Lande in 1981. Additive gene action is one of the major assumptions required for obtaining valid estimates. It is shown here that this assumption may be relaxed. One can estimate both a total number of effective loci and a number of dominant loci (the latter only when the parents are inbred) by comparing the variances of the F1, F2 and backcrosses. Numerical illustrations are given, based on crossbreeding data.

Bayesian statistical analyses for presence of single genes affecting meat quality traits in a crossed pig population.

Presence of single genes affecting meat quality traits was investigated in F2 individuals of a cross between Chinese Meishan and Western pig lines using phenotypic measurements on 11 traits. A Bayesian approach was used for inference about a mixed model of inheritance, postulating effects of polygenic background genes, action of a biallelic autosomal single gene and various nongenetic effects. Cooking loss, drip loss, two pH measurements, intramuscular fat, shearforce and back-fat thickness were traits found to be likely influenced by a single gene. In all cases, a recessive allele was found, which likely originates from the Meishan breed and is absent in the Western founder lines. By studying associations between genotypes assigned-to individuals based on phenotypic measurements for various traits, it was concluded that cooking loss, two pH measurements and possibly backfat thickness are influenced by one gene, and that a second gene influences intramuscular fat and possibly shearforce and drip loss. Statistical findings were supported by demonstrating marked differences in variances of families of fathers inferred as carriers and those inferred as noncarriers. It is concluded that further molecular genetic research effort to map single genes affecting these traits based on the same experimental data has a high probability of success.

Detection of quantitative trait loci for backfat thickness and intramuscular fat content in pigs (Sus scrofa).

In an experimental cross between Meishan and Dutch Large White and Landrace lines, 619 F(2) animals and their parents were typed for molecular markers covering the entire porcine genome. Associations were studied between these markers and two fatness traits: intramuscular fat content and backfat thickness. Association analyses were performed using interval mapping by regression under two genetic models: (1) an outbred line-cross model where the founder lines were assumed to be fixed for different QTL alleles; and (2) a half-sib model where a unique allele substitution effect was fitted within each of the 19 half-sib families. Both approaches revealed for backfat thickness a highly significant QTL on chromosome 7 and suggestive evidence for a QTL at chromosome 2. Furthermore, suggestive QTL affecting backfat thickness were detected on chromosomes 1 and 6 under the line-cross model. For intramuscular fat content the line-cross approach showed suggestive evidence for QTL on chromosomes 2, 4, and 6, whereas the half-sib analysis showed suggestive linkage for chromosomes 4 and 7. The nature of the QTL effects and assumptions underlying both models could explain discrepancies between the findings under the two models. It is concluded that both approaches can complement each other in the analysis of data from outbred line crosses.

SNP annotation-based whole genomic prediction and selection: an application to feed efficiency and its component traits in pigs.

The study investigated genetic architecture and predictive ability using genomic annotation of residual feed intake (RFI) and its component traits (daily feed intake [DFI], ADG, and back fat [BF]). A total of 1,272 Duroc pigs had both genotypic and phenotypic records, and the records were split into a training (968 pigs) and a validation dataset (304 pigs) by assigning records as before and after January 1, 2012, respectively. SNP were annotated by 14 different classes using Ensembl variant effect prediction. Predictive accuracy and prediction bias were calculated using Bayesian Power LASSO, Bayesian A, B, and Cπ, and genomic BLUP (GBLUP) methods. Predictive accuracy ranged from 0.508 to 0.531, 0.506 to 0.532, 0.276 to 0.357, and 0.308 to 0.362 for DFI, RFI, ADG, and BF, respectively. BayesCπ100.1 increased accuracy slightly compared to the GBLUP model and other methods. The contribution per SNP to total genomic variance was similar among annotated classes across different traits. Predictive performance of SNP classes did not significantly differ from randomized SNP groups. Genomic prediction has accuracy comparable to observed phenotype, and use of genomic prediction can be cost effective by replacing feed intake measurement. Genomic annotation had less impact on predictive accuracy traits considered here but may be different for other traits. It is the first study to provide useful insights into biological classes of SNP driving the whole genomic prediction for complex traits in pigs.

Heritabilities of and genetic relationships between salmonella resistance traits in broilers.

Using experimental infections, three traits for salmonella resistance were studied: mortality, survival time (in animals that died by infection), and quantitative cecal salmonella carriage at the end of the rearing period (in animals that did not die). In total, 548 animals were used; mortality was 29.2%, mean survival time was 5.97 d (n = 160), and the mean 10log of colony forming units per gram of cecal contents was 1.62 (n = 387). Genetic parameters were evaluated in bivariate threshold-linear models to account for the selective measurement of survival time and cecal carriage. Heritabilities were .06 for survival time, .09 for cecal carriage, and .12 for mortality. The genetic correlation between mortality and cecal carriage was weak (.26), which suggests that these traits are largely controlled by different genes. The genetic correlation between mortality and survival time was relatively strong (-.68). Simultaneous study of multiple traits seems to be of particular importance in judging epidemiological consequences of a possible selection for resistance. Results here indicate that selection on decreased mortality could be unfavorable for the spread of salmonella because the resulting correlated increase in survival time, implying longer shedding by infected animals, is relatively stronger than the correlated decrease in level of cecal carriage. Selection to reduce the level of cecal salmonella carriage could be done while keeping survival time constant, if so desired, because the correlation between these traits is weak (-.15).

Segregation analyses for presence of major genes affecting growth, backfat, and litter size in Dutch Meishan crossbreds.

Presence of major genes was investigated for two growth traits, backfat thickness, and two litter size traits in the F1 and F2 population of a cross between Meishan and European "White" pig lines. Segregation analyses were performed in a Bayesian setting, estimating the contribution of background polygenes and the contribution of a possible major gene to the expression of the traits considered. In a first analysis, F1 and F2 crossbred data were evaluated; different error variances were fitted for F1 and F2 observations. In the first analysis, significant contributions of major-gene variance were found for the two growth traits, backfat thickness, and litter size at first parity. In a second analysis, F2 data were evaluated to determine whether biases were introduced in the joint analysis of F1 and F2 data. In the second analysis, no major genes were found for growth traits. Major genes affecting backfat and litter size at first parity were confirmed. The gene identified to affect backfat is a dominant gene; the homozygous recessive genotype has approximately 6 mm of additional backfat. The gene identified to affect litter size at first parity also is a dominant gene; the homozygous recessive genotype produces five to six fewer pigs per litter.

Disease-induced variability of genetic correlations: ascites in broilers as a case study.

Breeding against a production disease is complicated by multiple relationships between productivity, disease, and environment. Ascites in broilers is such a disease. The combination of the reasonably well understood etiology (a physiological/pathological cascade due to inadequate oxygen supply) and the practical relevance makes ascites a relevant case for demonstrating and partly revealing these complex relationships. Chickens (n = 2,788) were tested in an ascites-challenging (cold) environment. Genetic analysis of mortality and pathology in combination with performance and physiological traits (especially blood gas traits) revealed ample opportunities for selection against ascites expression. The genetic correlation structure indicated that different mortality traits and pathology traits roughly represent one common characteristic. Direct selection against pathology is more effective than selection on the basis of growth or blood gas traits. The observed negative correlation (-0.26) between productivity and ascites was unexpected. From the etiology of ascites (inadequate supply of oxygen relative to the demand), a positive (unfavorable) correlation was expected. To demonstrate that the actual disease occurrence caused this apparent contradiction, the data from the undiseased subpopulation were reanalyzed. In the undiseased subpopulation, the genetic correlation between productivity and ascites was positive (0.29). This discrepancy was confirmed by comparing regression of ascites expression on actual performance with regression of ascites on independently assessed performance breeding values. The lability of the genetic correlation was explained from complex interactions between productivity, disease susceptibility, and actual occurrence of the disease. The revealed mechanism can be generalized to other production-related diseases and results in systematically lower genetic correlations between disease and productivity. It was inferred that genetic correlations between productivity and such diseases will always be prone to the demonstrated environmental sensitivity, which complicates index selection against production-related diseases.