Limits...
Incorporating parent-of-origin effects in whole-genome prediction of complex traits.

Hu Y, Rosa GJ, Gianola D - Genet. Sel. Evol. (2016)

Bottom Line: The simulation and the negative result obtained in the real data analysis indicated that, in order to gain benefit from the POE model in terms of prediction, a sizable contribution of parent-of-origin effects to variation is needed and such variation must be captured by the genetic markers fitted.Recent studies, however, suggest that most parent-of-origin effects stem from epigenetic regulation but not from a change in DNA sequence.Therefore, integrating epigenetic information with genetic markers may help to account for parent-of-origin effects in whole-genome prediction.

View Article: PubMed Central - PubMed

Affiliation: Department of Animal Sciences, University of Wisconsin-Madison, 1675 Observatory Dr., Madison, WI, 53706, USA. yhu32@wisc.edu.

ABSTRACT

Background: Parent-of-origin effects are due to differential contributions of paternal and maternal lineages to offspring phenotypes. Such effects include, for example, maternal effects in several species. However, epigenetically induced parent-of-origin effects have recently attracted attention due to their potential impact on variation of complex traits. Given that prediction of genetic merit or phenotypic performance is of interest in the study of complex traits, it is relevant to consider parent-of-origin effects in such predictions. We built a whole-genome prediction model that incorporates parent-of-origin effects by considering parental allele substitution effects of single nucleotide polymorphisms and gametic relationships derived from a pedigree (the POE model). We used this model to predict body mass index in a mouse population, a trait that is presumably affected by parent-of-origin effects, and also compared the prediction performance to that of a standard additive model that ignores parent-of-origin effects (the ADD model). We also used simulated data to assess the predictive performance of the POE model under various circumstances, in which parent-of-origin effects were generated by mimicking an imprinting mechanism.

Results: The POE model did not predict better than the ADD model in the real data analysis, probably due to overfitting, since the POE model had far more parameters than the ADD model. However, when applied to simulated data, the POE model outperformed the ADD model when the contribution of parent-of-origin effects to phenotypic variation increased. The superiority of the POE model over the ADD model was up to 8 % on predictive correlation and 5 % on predictive mean squared error.

Conclusions: The simulation and the negative result obtained in the real data analysis indicated that, in order to gain benefit from the POE model in terms of prediction, a sizable contribution of parent-of-origin effects to variation is needed and such variation must be captured by the genetic markers fitted. Recent studies, however, suggest that most parent-of-origin effects stem from epigenetic regulation but not from a change in DNA sequence. Therefore, integrating epigenetic information with genetic markers may help to account for parent-of-origin effects in whole-genome prediction.

Show MeSH
Trend of averaged predictive correlation and MSE with change of s (proportion of imprinted QTL) under  (complete imprinting). Predictive correlation and MSE decrease as s goes up for both models. ADD additive model, POE parent-of-origin effects model
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4834899&req=5

Fig3: Trend of averaged predictive correlation and MSE with change of s (proportion of imprinted QTL) under (complete imprinting). Predictive correlation and MSE decrease as s goes up for both models. ADD additive model, POE parent-of-origin effects model

Mentions: When imprinting was complete (), it was not surprising that the POE model performed better than the additive ADD model. The superiority of the POE model over the ADD model depended on s, i.e., the larger the proportion of imprinted genes, the bigger the difference on predictive correlation and MSE between the two models. As s increased, a larger fraction of genetic variation was attributed to parent-of-origin effects, which cannot be captured by the ADD model. An interesting observation from Figs. 1 and 2 is that, for a given model, the predictive correlation and MSE decreased with an increase of s (Fig. 3). Recall that the data was simulated such that the allele effect was multiplied by (less imprinting as ), and fraction s of all QTL were assumed to be imprinted (Eq. 8). Suppose a QTL is maternally imprinted (the allele inherited from the mother written first), and that the values of the four genotypes (expressed as deviations from the population mean) are:9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned} \begin{aligned} G_{11}&= \rho \cdot \theta _1 + \theta _1,\\ G_{21}&= \rho \cdot \theta _2 + \theta _1,\\ G_{12}&= \rho \cdot \theta _1 + \theta _2,\\ G_{22}&= \rho \cdot \theta _2 + \theta _2. \end{aligned} \end{aligned}$$\end{document}G11=ρ·θ1+θ1,G21=ρ·θ2+θ1,G12=ρ·θ1+θ2,G22=ρ·θ2+θ2.Let p and q be the frequencies of the and alleles. The genetic variance at this locus can be calculated as:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned} \sigma ^2&= P_{11}\cdot G^2_{11} + P_{21}\cdot G^2_{21} + P_{12}\cdot G^2_{12} + P_{22}\cdot G^2_{22} \\&=(1+\rho ^2)pq(\theta _1-\theta _2)^2, \end{aligned}$$\end{document}σ2=P11·G112+P21·G212+P12·G122+P22·G222=(1+ρ2)pq(θ1-θ2)2,where is the genotype frequency of assuming Hardy-Weinberg equilibrium. Note that is , the allele substitution effect defined by a standard additive genetic model. From Eq. 10, when (no imprinting), the expression yields , the additive variance of a standard genetic model (e.g., [91, 92]). When , however, this variance (“signal”) decreases as approaches 0 (i.e., increased imprinting level). Hence, for a given value of that is smaller than 1 (0 in this case), the total variance of all QTL becomes smaller as s increases. Since the environmental distribution was the same in all settings, heritability decreased as s increased, producing a lower predictive ability.Fig. 3


Incorporating parent-of-origin effects in whole-genome prediction of complex traits.

Hu Y, Rosa GJ, Gianola D - Genet. Sel. Evol. (2016)

Trend of averaged predictive correlation and MSE with change of s (proportion of imprinted QTL) under  (complete imprinting). Predictive correlation and MSE decrease as s goes up for both models. ADD additive model, POE parent-of-origin effects model
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4834899&req=5

Fig3: Trend of averaged predictive correlation and MSE with change of s (proportion of imprinted QTL) under (complete imprinting). Predictive correlation and MSE decrease as s goes up for both models. ADD additive model, POE parent-of-origin effects model
Mentions: When imprinting was complete (), it was not surprising that the POE model performed better than the additive ADD model. The superiority of the POE model over the ADD model depended on s, i.e., the larger the proportion of imprinted genes, the bigger the difference on predictive correlation and MSE between the two models. As s increased, a larger fraction of genetic variation was attributed to parent-of-origin effects, which cannot be captured by the ADD model. An interesting observation from Figs. 1 and 2 is that, for a given model, the predictive correlation and MSE decreased with an increase of s (Fig. 3). Recall that the data was simulated such that the allele effect was multiplied by (less imprinting as ), and fraction s of all QTL were assumed to be imprinted (Eq. 8). Suppose a QTL is maternally imprinted (the allele inherited from the mother written first), and that the values of the four genotypes (expressed as deviations from the population mean) are:9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned} \begin{aligned} G_{11}&= \rho \cdot \theta _1 + \theta _1,\\ G_{21}&= \rho \cdot \theta _2 + \theta _1,\\ G_{12}&= \rho \cdot \theta _1 + \theta _2,\\ G_{22}&= \rho \cdot \theta _2 + \theta _2. \end{aligned} \end{aligned}$$\end{document}G11=ρ·θ1+θ1,G21=ρ·θ2+θ1,G12=ρ·θ1+θ2,G22=ρ·θ2+θ2.Let p and q be the frequencies of the and alleles. The genetic variance at this locus can be calculated as:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned} \sigma ^2&= P_{11}\cdot G^2_{11} + P_{21}\cdot G^2_{21} + P_{12}\cdot G^2_{12} + P_{22}\cdot G^2_{22} \\&=(1+\rho ^2)pq(\theta _1-\theta _2)^2, \end{aligned}$$\end{document}σ2=P11·G112+P21·G212+P12·G122+P22·G222=(1+ρ2)pq(θ1-θ2)2,where is the genotype frequency of assuming Hardy-Weinberg equilibrium. Note that is , the allele substitution effect defined by a standard additive genetic model. From Eq. 10, when (no imprinting), the expression yields , the additive variance of a standard genetic model (e.g., [91, 92]). When , however, this variance (“signal”) decreases as approaches 0 (i.e., increased imprinting level). Hence, for a given value of that is smaller than 1 (0 in this case), the total variance of all QTL becomes smaller as s increases. Since the environmental distribution was the same in all settings, heritability decreased as s increased, producing a lower predictive ability.Fig. 3

Bottom Line: The simulation and the negative result obtained in the real data analysis indicated that, in order to gain benefit from the POE model in terms of prediction, a sizable contribution of parent-of-origin effects to variation is needed and such variation must be captured by the genetic markers fitted.Recent studies, however, suggest that most parent-of-origin effects stem from epigenetic regulation but not from a change in DNA sequence.Therefore, integrating epigenetic information with genetic markers may help to account for parent-of-origin effects in whole-genome prediction.

View Article: PubMed Central - PubMed

Affiliation: Department of Animal Sciences, University of Wisconsin-Madison, 1675 Observatory Dr., Madison, WI, 53706, USA. yhu32@wisc.edu.

ABSTRACT

Background: Parent-of-origin effects are due to differential contributions of paternal and maternal lineages to offspring phenotypes. Such effects include, for example, maternal effects in several species. However, epigenetically induced parent-of-origin effects have recently attracted attention due to their potential impact on variation of complex traits. Given that prediction of genetic merit or phenotypic performance is of interest in the study of complex traits, it is relevant to consider parent-of-origin effects in such predictions. We built a whole-genome prediction model that incorporates parent-of-origin effects by considering parental allele substitution effects of single nucleotide polymorphisms and gametic relationships derived from a pedigree (the POE model). We used this model to predict body mass index in a mouse population, a trait that is presumably affected by parent-of-origin effects, and also compared the prediction performance to that of a standard additive model that ignores parent-of-origin effects (the ADD model). We also used simulated data to assess the predictive performance of the POE model under various circumstances, in which parent-of-origin effects were generated by mimicking an imprinting mechanism.

Results: The POE model did not predict better than the ADD model in the real data analysis, probably due to overfitting, since the POE model had far more parameters than the ADD model. However, when applied to simulated data, the POE model outperformed the ADD model when the contribution of parent-of-origin effects to phenotypic variation increased. The superiority of the POE model over the ADD model was up to 8 % on predictive correlation and 5 % on predictive mean squared error.

Conclusions: The simulation and the negative result obtained in the real data analysis indicated that, in order to gain benefit from the POE model in terms of prediction, a sizable contribution of parent-of-origin effects to variation is needed and such variation must be captured by the genetic markers fitted. Recent studies, however, suggest that most parent-of-origin effects stem from epigenetic regulation but not from a change in DNA sequence. Therefore, integrating epigenetic information with genetic markers may help to account for parent-of-origin effects in whole-genome prediction.

Show MeSH