Levels Of Genetic Correlation Research Articles

Comparative Study of Single-Trait and Multi-Trait Genomic Prediction Models.

Conventional genomic selection models trait individually, neglecting complex trait interactions. Multi-trait models address this by considering genetic correlations, thus improving breeding value accuracy. Despite their theoretical benefits, quantifying these models' breeding advantages across genetic backgrounds is essential. This study evaluates the benefits of multi-trait models under varying population sizes and three levels of genetic correlations (low, medium, high) using simulations based on 50 K chip data from 5000 individuals. In equal heritability scenarios, the multi-trait GBLUP model consistently outperforms single-trait models, with breeding advantages increasing with heritability. For example, with a reference population of 4500, improvements range from 0.3% to 4.1%. Notably, trait combinations with low heritability are insensitive to changes in genetic correlation, with gains remaining ≤ 0.1% across different genetic correlations under low heritability conditions. In differing heritability scenarios, the multi-trait model's benefits vary, particularly enhancing low-heritability traits when paired with high-heritability ones. Additionally, modeling time increases as genetic correlation decreases. The results of this study indicate that multi-trait models improve breeding accuracy but require more modeling time and place higher demands on algorithms and software. We recommend breeding strategies tailored to different phenotypes and genetic backgrounds to balance efficiency and accuracy.

GbyE: an integrated tool for genome widely association study and genome selection based on genetic by environmental interaction.

The growth and development of organism were dependent on the effect of genetic, environment, and their interaction. In recent decades, lots of candidate additive genetic markers and genes had been detected by using genome-widely association study (GWAS). However, restricted to computing power and practical tool, the interactive effect of markers and genes were not revealed clearly. And utilization of these interactive markers is difficult in the breeding and prediction, such as genome selection (GS). Through the Power-FDR curve, the GbyE algorithm can detect more significant genetic loci at different levels of genetic correlation and heritability, especially at low heritability levels. The additive effect of GbyE exhibits high significance on certain chromosomes, while the interactive effect detects more significant sites on other chromosomes, which were not detected in the first two parts. In prediction accuracy testing, in most cases of heritability and genetic correlation, the majority of prediction accuracy of GbyE is significantly higher than that of the mean method, regardless of whether the rrBLUP model or BGLR model is used for statistics. The GbyE algorithm improves the prediction accuracy of the three Bayesian models BRR, BayesA, and BayesLASSO using information from genetic by environmental interaction (G × E) and increases the prediction accuracy by 9.4%, 9.1%, and 11%, respectively, relative to the Mean value method. The GbyE algorithm is significantly superior to the mean method in the absence of a single environment, regardless of the combination of heritability and genetic correlation, especially in the case of high genetic correlation and heritability. Therefore, this study constructed a new genotype design model program (GbyE) for GWAS and GS using Kronecker product. which was able to clearly estimate the additive and interactive effects separately. The results showed that GbyE can provide higher statistical power for the GWAS and more prediction accuracy of the GS models. In addition, GbyE gives varying degrees of improvement of prediction accuracy in three Bayesian models (BRR, BayesA, and BayesCpi). Whatever the phenotype were missed in the single environment or multiple environments, the GbyE also makes better prediction for inference population set. This study helps us understand the interactive relationship between genomic and environment in the complex traits. The GbyE source code is available at the GitHub website ( https://github.com/liu-xinrui/GbyE ).

Modeling Illustrates That Genomic Selection Provides New Opportunities for Intercrop Breeding.

Intercrop breeding programs using genomic selection can produce faster genetic gain than intercrop breeding programs using phenotypic selection. Intercropping is an agricultural practice in which two or more component crops are grown together. It can lead to enhanced soil structure and fertility, improved weed suppression, and better control of pests and diseases. Especially in subsistence agriculture, intercropping has great potential to optimize farming and increase profitability. However, breeding for intercrop varieties is complex as it requires simultaneous improvement of two or more component crops that combine well in the field. We hypothesize that genomic selection can significantly simplify and accelerate the process of breeding crops for intercropping. Therefore, we used stochastic simulation to compare four different intercrop breeding programs implementing genomic selection and an intercrop breeding program entirely based on phenotypic selection. We assumed three different levels of genetic correlation between monocrop grain yield and intercrop grain yield to investigate how the different breeding strategies are impacted by this factor. We found that all four simulated breeding programs using genomic selection produced significantly more intercrop genetic gain than the phenotypic selection program regardless of the genetic correlation with monocrop yield. We suggest a genomic selection strategy which combines monocrop and intercrop trait information to predict general intercropping ability to increase selection accuracy in the early stages of a breeding program and to minimize the generation interval.

A deterministic equation to predict the accuracy of multi-population genomic prediction with multiple genomic relationship matrices

BackgroundA multi-population genomic prediction (GP) model in which important pre-selected single nucleotide polymorphisms (SNPs) are differentially weighted (MPMG) has been shown to result in better prediction accuracy than a multi-population, single genomic relationship matrix ({mathbf{GRM}}) GP model (MPSG) in which all SNPs are weighted equally. Our objective was to underpin theoretically the advantages and limits of the MPMG model over the MPSG model, by deriving and validating a deterministic prediction equation for its accuracy.MethodsUsing selection index theory, we derived an equation to predict the accuracy of estimated total genomic values of selection candidates from population A (r_{{{mathbf{EGV}}_{{A_{T} }} }}), when individuals from two populations, A and B, are combined in the training population and two {mathbf{GRM}}, made respectively from pre-selected and remaining SNPs, are fitted simultaneously in MPMG. We used simulations to validate the prediction equation in scenarios that differed in the level of genetic correlation between populations, heritability, and proportion of genetic variance explained by the pre-selected SNPs. Empirical accuracy of the MPMG model in each scenario was calculated and compared to the predicted accuracy from the equation.ResultsIn general, the derived prediction equation resulted in accurate predictions of r_{{{mathbf{EGV}}_{{A_{T} }} }} for the scenarios evaluated. Using the prediction equation, we showed that an important advantage of the MPMG model over the MPSG model is its ability to benefit from the small number of independent chromosome segments (M_{e}) due to the pre-selected SNPs, both within and across populations, whereas for the MPSG model, there is only a single value for M_{e}, calculated based on all SNPs, which is very large. However, this advantage is dependent on the pre-selected SNPs that explain some proportion of the total genetic variance for the trait.ConclusionsWe developed an equation that gives insight into why, and under which conditions the MPMG outperforms the MPSG model for GP. The equation can be used as a deterministic tool to assess the potential benefit of combining information from different populations, e.g., different breeds or lines for GP in livestock or plants, or different groups of people based on their ethnic background for prediction of disease risk scores.

Genetic parameters of backfat fatty acids and carcass traits in Large White pigs

Subcutaneous fat thickness and fatty acid composition (FAC) play an important role on seasoning loss and organoleptic characteristics of seasoned hams. Dry-cured ham industry prefers meats with low contents of polyunsaturated fatty acids (PUFA) because these negatively affect fat firmness and ham quality, whereas consumers require higher contents in those fatty acids (FA) for their positive effect on human health. A population of 950 Italian Large White pigs from the Italian National Sib Test Selection Programme was investigated with the aim to estimate heritabilities, genetic and phenotypic correlations of backfat FAC, Semimembranosus muscle intramuscular fat (IMF) content and other carcass traits. The pigs were reared in controlled environmental condition at the same central testing station and were slaughtered at reaching 150 kg live weight. Backfat samples were collected to analyze FAC by gas chromatography. Carcass traits showed heritability levels from 0.087 for estimated carcass lean percentage to 0.361 for hot carcass weight. Heritability values of FA classes were low-to-moderate, all in the range 0.245 for n-3 PUFA to 0.264 for monounsaturated FA (MUFA). Polyunsaturated fatty acids showed a significant genetic correlation with loin thickness (0.128), backfat thickness (−0.124 for backfat measured by Fat-O-Meat’er and −0.175 for backfat measured by calibre) and IMF (−0.102). Obviously, C18:2(n-6) shows similar genetic correlations with the same traits (0.211 with loin thickness, −0.206 with backfat measured by Fat-O-Meat’er, −0.291 with backfat measured by calibre and −0.171 with IMF). Monounsaturated FA, except with the backfat measured by calibre (0.068; P<0.01), do not show genetic correlations with carcass characteristics, whereas a negative genetic correlation was found between MUFA and saturated FA (SFA; −0.339; P<0.001). These results suggest that MUFA/SFA ratio could be increased without interfering with carcass traits. The level of genetic correlations between FA and carcass traits should be taken into account in dealing with the development of selection schemes addressed to modify carcass composition and/or backfat FAC.

Comparison of repeatability and multiple traits model in estimating heritability, breeding values and genetic correlations for four continuous traits using different criteria of evaluation (a simulation study).

ABSTRACTThe present study evaluated the multiple traits model, as opposite to repeatability model approach in estimating heritability, breeding values and genetic correlation for four continuous traits (represent four parities) using bias, Mean squared error (MSE), Akaike information criteria (AIC) and Bayesian information criteria (BIC). The simulated base population consisted of 20, 60 or 100 sires. Each sire mated to 50 females to produce 1000, 3000 or 5000 progenies. The variance components modified to simulate three levels of heritability (h2), 0.05, 0.25 and 0.5. Genetic correlations (GC) among the four studied traits were 0.3, 0.5, 0.9 and one and residual correlation was 0.2. Twenty replicates were generated for each of the 36 combinations (three levels of h2 * four levels of GC * three levels of number of animals). These data sets (36 set) divided into two scenarios. The first one (27 set) was simulated as multiple traits and the second scenario (nine sets) was simulated as repeated measures. Each set of data in each scenario analyzed by multiple traits model (MTM) and repeatability model (RM). Correlations between the true and estimated breeding values of studied trait(s) estimated for the two types of analysis. Then, bias, MSE, AIC and BIC for all estimated values calculated as measures for comparing models of estimation. The mean estimates of h2 resulted from MTM and RM were 0.27, and 0.20, respectively. When level of GC increases, the mean estimate of h2 increases and reaches to equal the estimate of h2 resulted from MTM when GC=1. The bias and MSE of MTM are less than those of RM. The smallest estimates of bias and MSE were noticed at GC = 0.9 and 1. The lowest AIC and BIC values were observed when fitting RM with data of the two scenarios. Therefore, the two criteria favored RM. Correlations between true and estimated breeding values of the four traits in MTM were slightly better than in RM. The effect of type of model was significant (P<0.01). In addition, significant effects between number of animals levels, heritability levels and GC levels were observed (P<0.01). This study indicated that, multiple traits analysis is more accurate than repeated measurements analysis in estimating h2 and breeding values as concluded from results of bias and MSE. AIC and BIC were not the suitable criteria for selecting the appropriate model under the circumstances of this study.

Incorporation of economic values into the component traits of a ratio: Feed efficiency

Direct selection on a ratio (R) of 2 traits (x1/x2) does not have a mechanism to accommodate the relative economic values (a1 and a2) between x1 and x2 because selection criteria x1/x2 and a1x1/a2x2 rank animals in the same order. This study presented a procedure to incorporate the economic weights into ratio traits through linear transformation. The partial derivatives of a nonlinear profit function evaluated at the means were widely taken as economic weights in the literature. This study showed that the economic weights derived in this manner were erroneous because they actually contain a mixture of actual economic weights and transformation effects. The ratios 1/2 and 2/4 are considered equal by selection on R, but are treated differently by the linear index. In addition, this study presented a unified approach to compare 4 different selection strategies for genetic improvement of ratio traits: linear index (I), selection on the ratio (R), selection on difference between x1 and x2 (D), and selection on x1 alone. This study considered 3 levels of heritability each for variables x1 and x2 and 2 levels of genetic correlations (γG), 2 ratios of means (µ1/µ2), and 4 ratios of phenotypic variances giving a total of 96 scenarios. Linear index I was the most efficient of the 4 criteria compared in all 96 scenarios studied. The superiority of index I over R, D, and selection on x1 alone are particularly remarkable when x1 and x2 have a large difference in heritability and are highly correlated. Selection on x1 alone is an economically viable alternative to criterion I or R for the improvement of ratio traits particularly when x1 is more heritable than x2 and when x2 is costly to measure. Selection on D is more efficient than direct selection on R or selection on x1 alone when x1 is less heritable than x2 and the difference between µ1 and µ2 is small.

Applications of multi-trait selection in common bean using real and simulated experiments

The analysis of multiples traits has been largely employed in animal and forestry crop breeding; however, in annual crops such as common bean this approach is practically lacking. This study examined the selective accuracy and estimates of genetic parameters in common bean using single- and multi-trait analyses in several designs of genetic correlations and heritabilities by means of simulated- and real-traits. We also compared the efficiency of the Bayesian and REML methods as regarding its selective accuracy and estimates of genetic parameters. In real data experiments, 100 endogamic families of common bean were assessed for grain yield (GY) and grain type (GT) in three environments using a 10 × 10 triple lattice design. Single- and multi-trait analyses were applied for these traits using REML-based BLUP analysis and Bayesian methods. In simulated experiments, traits corresponding to GY and yield components in common bean were simulated under several levels of genetic correlations and evaluated under different levels of heritability in order to verify in which genetic/experimental design it is advantageous to apply single- or multi-trait analyses. The results obtained in this study indicated that GY and GT present low genetic correlation and, therefore, multi-trait analysis may be unjustifiable for these traits. The Bayesian and REML multi-trait analyses were equivalent in point and interval estimates of genetic parameters and predictive accuracy, both for the field and simulated experiments. In addition, it was found that multi-trait analysis is justified in situations where the traits present low/moderate heritability and medium/high genetic correlations among them. It also was found that GY in common bean may be outstandingly favored in accuracy when evaluated jointly with its components. Thus, one can conclude that the simultaneous analysis of GY and yield components in common bean may provide superior gains accuracy and selection efficiency, being the response to selection better obtained by use of the multi-trait heritability.

Accuracy of multi-trait genomic selection using different methods

BackgroundGenomic selection has become a very important tool in animal genetics and is rapidly emerging in plant genetics. It holds the promise to be particularly beneficial to select for traits that are difficult or expensive to measure, such as traits that are measured in one environment and selected for in another environment. The objective of this paper was to develop three models that would permit multi-trait genomic selection by combining scarcely recorded traits with genetically correlated indicator traits, and to compare their performance to single-trait models, using simulated datasets.MethodsThree (SNP) Single Nucleotide Polymorphism based models were used. Model G and BCπ0 assumed that contributed (co)variances of all SNP are equal. Model BSSVS sampled SNP effects from a distribution with large (or small) effects to model SNP that are (or not) associated with a quantitative trait locus. For reasons of comparison, model A including pedigree but not SNP information was fitted as well.ResultsIn terms of accuracies for animals without phenotypes, the models generally ranked as follows: BSSVS > BCπ0 > G > > A. Using multi-trait SNP-based models, the accuracy for juvenile animals without any phenotypes increased up to 0.10. For animals with phenotypes on an indicator trait only, accuracy increased up to 0.03 and 0.14, for genetic correlations with the evaluated trait of 0.25 and 0.75, respectively.ConclusionsWhen the indicator trait had a genetic correlation lower than 0.5 with the trait of interest in our simulated data, the accuracy was higher if genotypes rather than phenotypes were obtained for the indicator trait. However, when genetic correlations were higher than 0.5, using an indicator trait led to higher accuracies for selection candidates. For different combinations of traits, the level of genetic correlation below which genotyping selection candidates is more effective than obtaining phenotypes for an indicator trait, needs to be derived considering at least the heritabilities and the numbers of animals recorded for the traits involved.

The efficiency of genome-wide selection for genetic improvement of net merit1

Four methods of selection for net merit comprising 2 correlated traits were compared in this study: 1) EBV-only index (I₁), which consists of the EBV of both traits (i.e., traditional 2-trait BLUP selection); 2) GEBV-only index (I₂), which comprises the genomic EBV (GEBV) of both traits; 3) GEBV-assisted index (I₃), which combines both the EBV and the GEBV of both traits; and 4) GBV-assisted index (I₄), which combines both the EBV and the true genomic breeding value (GBV) of both traits. Comparisons of these indices were based on 3 evaluation criteria [selection accuracy, genetic response (ΔH), and relative efficiency] under 64 scenarios that arise from combining 2 levels of genetic correlation (r(G)), 2 ratios of genetic variances between traits, 2 ratios of the genomic variance to total genetic variances for trait 1, 4 accuracies of EBV, and 2 proportions of r(G) explained by the GBV. Both selection accuracy and genetic responses of the indices I₁, I₃, and I₄ increased as the accuracy of EBV increased, but the efficiency of the indices I₃ and I₄ relative to I₁ decreased as the accuracy of EBV increased. The relative efficiency of both I₃ and I₄ was generally greater when the accuracy of EBV was 0.6 than when it was 0.9, suggesting that the genomic markers are most useful to assist selection when the accuracy of EBV is low. The GBV-assisted index I₄ was superior to the GEBV-assisted I₃ in all 64 cases examined, indicating the importance of improving the accuracy of prediction of genomic breeding values. Other parameters being identical, increasing the genetic variance of a high heritability trait would increase the genetic response of the genomic indices (I₂, I₃, and I₄). The genetic responses to I₂, I₃, and I(4) was greater when the genetic correlation between traits was positive (r(G) = 0.5) than when it was negative (r(G) = -0.5). The results of this study indicate that the effectiveness of the GEBV-assisted index I₃ is affected by heritability of and genetic correlation between traits, the ratio of genetic variances between traits, the genomic-genetic variance ratio of each index trait, the proportion of genetic correlation accounted for by the genomic markers, and the accuracy of predictions of both EBV and GBV. However, most of these affecting factors are genetic characteristics of a population that is beyond the control of the breeders. The key factor subject to manipulation is to maximize both the proportion of the genetic variance explained by GEBV and the accuracy of both GEBV and EBV. The developed procedures provide means to investigate the efficiency of various genomic indices for any given combination of the genetic factors studied.

Accuracy of Single- and Multiple-Trait REML Evaluation of Data Including Non-Random Missing Records

AbstractWe examined the accuracy of single- and multiple-trait REML procedures by studying estimates of within-individual genetic correlations between an ordered categorical trait and a continuous trait. The traits were derived from simulated bivariate, normally distributed data including selectively deleted records. Ten thousand data sets were generated for each partially factorial combination of two levels of genetic correlation (0.3 and 0.6), and environmental correlation (0.3 and 0.6), and three levels of narrow-sense individual heritability (0.05, 0.15 and 0.25) and mortality (0, 10, 30 and 50%). All data sets consisted of data on 200 unrelated parents, each with 20 halfsib progenies. The accuracy of the evaluations was illustrated in terms of average bias and variation of derived correlation estimates. The average bias values generated by multiple-trait REML were generally low. In contrast, single-trait REML was sensitive to selective deletion of records and systematically underestimated the genetic correlations. For both methods, especially at low heritabilities, the magnitude of the variation was generally high, showing that there is a substantial probability of obtaining seriously misleading genetic correlation estimates if the analysis is based on a single experiment and data include non-random missing records.

Efficiency of drained selection environments for improving grain yield in faba bean under undrained target environments on vertisol

Waterlogging is a major constraint of faba bean (Vicia faba L.) production in the highland vertisols of Ethiopia. Genotypes are currently evaluated under both the drained conditions of Broad Bed and Furrow(BBF) and the undrained flatbeds. The objective of this research was to determine if selection of genotypes under BBF is efficient for identification of genotypes for the flatbed conditions. Yields under the flatbed conditions were lower by 20–50% (38% on average) than yields under the BBF. Genetic variation, heritability and predicted genetic gains for grain yield due to direct selection were also lower under the flatbeds than they are under the BBF. There were good levels of genetic correlation (r = 0.20–0.75 or0.45 on average) between yield performances under BBF and flatbeds, indicating that selection under BBF may identify good genotypes for flatbeds and vice versa. There was also high association (r =0.91) between relative yield reduction under flatbeds and the ratio of the predicted genetic gains from indirect selection under BBF to the predicted genetic gains from direct selection under flat beds. This indicates that indirect selection under BBF may be useful to identify better genotypes for flatbed conditions, but the relative efficiency of indirect selection was lower than direct selection under flatbeds themselves.

Lexical and grammatical development: a behavioural genetic perspective.

The relation of lexical and grammatical knowledge is at the core of many controversies in linguistics and psycholinguistics. Recent empirical findings that the two are highly correlated in early language development have further energized the theoretical debate. Behavioural genetics provides an illuminating new tool to explore this question, by addressing the question of whether the empirical correlation simply reflects the fact that environments which facilitate one aspect of language growth also facilitate the other, or whether the same underlying acquisition mechanisms, influenced by the same genes, are responsible for the correlation. We explored this issue in a study of 2898 pairs of two-year-old twins born in England and Wales. Language development was assessed by their parents using an adapted version of the MacArthur Communicative Development Inventory which assesses vocabulary and grammar. Moderate heritabilities were found for both. As in previous studies, measures of vocabulary and sentence complexity were substantially correlated (r = 0.66). Behaviour-genetic modelling of the relation of vocabulary and grammar produced an estimated value of 0.61 for the genetic correlation, a measure of the overlap of the genetic effects that contribute to the two aspects of language development. In contrast, a measure of nonverbal cognitive development, the PARCA, was only weakly correlated at both the phenotypic level and at the level of genetic correlations with the language measures. Thus, although the distinction between verbal and nonverbal skills has a genetic basis underlying the phenotypic dissociation, there is little evidence either genetically or phenotypically for a dissociation between vocabulary and grammar within language.