MegaLMM: Mega-scale linear mixed models for genomic predictions with thousands of traits

Runcie, Daniel E.; Qu, Jiayi; Cheng, Hao; Crawford, Lorin

doi:10.1186/s13059-021-02416-w

Cited by 47 publications

(52 citation statements)

References 84 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…, genetic and error covariances), which affects the prediction performance of the MT models as well as the accuracy of breeding value estimates. The larger the number of traits, the larger the required number of parameters that need to be estimated ( Runcie et al 2021 ). Also, the more complex the model is and the larger the number of traits included, the greater chances there are of facing convergence problems in the analysis ( Runcie et al 2021 ).…”

Section: Introductionmentioning

confidence: 99%

“…The larger the number of traits, the larger the required number of parameters that need to be estimated ( Runcie et al 2021 ). Also, the more complex the model is and the larger the number of traits included, the greater chances there are of facing convergence problems in the analysis ( Runcie et al 2021 ). This means that MT models require more data to be able to accurately estimate the additional parameters ( Okeke et al 2017 ).…”

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…Furthermore, the implementation of conventional MT (frequentist and Bayesian) models is, in general, computationally demanding ( Runcie et al 2021 ). The fragility of these methods is due to the number of variance–covariance parameters that must be estimated, which increases quadratically with the number of traits ( Runcie et al 2021 ). The computational demands increase even more dramatically, from cubically to quantically, with the number of traits ( Zhou and Stephens 2014 ) because most algorithms require repeated inversion of large covariance matrices.…”

Section: Introductionmentioning

confidence: 99%

“…The computational demands increase even more dramatically, from cubically to quantically, with the number of traits ( Zhou and Stephens 2014 ) because most algorithms require repeated inversion of large covariance matrices. These matrix operations dominate the time required to fit conventional MT models, leading to models that take days, weeks, or even years to converge ( Runcie et al 2021 ).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Bayesian multitrait kernel methods improve multienvironment genome-based prediction

Montesinos‐López

Montesinos-López

Montesinos‐López³

et al. 2021

G3 Genes|Genomes|Genetics

View full text Add to dashboard Cite

When multi-trait data are available, the preferred models are those that are able to account for correlations between phenotypic traits because when the degree of correlation is moderate or large, this increases the genomic prediction accuracy. For this reason, in this paper we explore Bayesian multi-trait kernel methods for genomic prediction and we illustrate the power of these models with three real datasets. The kernels under study were the linear, Gaussian, polynomial and sigmoid kernels; they were compared with the conventional Ridge regression and GBLUP multi-trait models. The results show that, in general, the Gaussian kernel method outperformed conventional Bayesian Ridge and GBLUP multi-trait linear models by 2.2 to 17.45% (datasets 1 to 3) in terms of prediction performance based on the mean square error of prediction. This improvement in terms of prediction performance of the Bayesian multi-trait kernel method can be attributed to the fact that the proposed model is able to capture non-linear patterns more efficiently than linear multi-trait models. However, not all kernels perform well in the datasets used for evaluation, which is why more than one kernel should be evaluated to be able to choose the best kernel.

show abstract

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Bayesian multitrait kernel methods improve multienvironment genome-based prediction

Montesinos‐López

Montesinos-López

Montesinos‐López³

et al. 2021

G3 Genes|Genomes|Genetics

View full text Add to dashboard Cite

show abstract

Predicting Genotype × Environment × Management (G × E × M) Interactions for the Design of Crop Improvement Strategies

Cooper

Messina

Tang

et al. 2022

Plant Breeding Reviews

View full text Add to dashboard Cite

Genomic prediction of metabolic content in rice grain in response to warmer night conditions

Bi,

Walia,

Obata

et al. 2024

Crop Science

View full text Add to dashboard Cite

It has been argued that metabolic content can be used as a selection marker to accelerate crop improvement because metabolic profiles in crops are often under genetic control. Evaluating the role of genetics in metabolic variation is a long‐standing challenge. Rice, one of the world's most important staple crops, is known to be sensitive to recent increases in nighttime temperatures. Quantification of metabolic levels can help measure rice responses to high night temperature (HNT) stress. However, the extent of metabolic variation that can be explained by regression on whole‐genome molecular markers remains to be evaluated. In the current study, we generated metabolic profiles for mature grains from a subset of rice diversity panel accessions grown under optimal and HNT conditions. Metabolite accumulation was low to moderately heritable, and genomic prediction accuracies of metabolite accumulation were within the expected upper limit set by their genomic heritability estimates. Genomic heritability estimates were slightly higher in the control group than in the HNT group. Genomic correlation estimates for the same metabolite accumulation between the control and HNT conditions indicated the presence of genotype‐by‐environment interactions. Reproducing kernel Hilbert spaces regression and image‐based deep learning improved prediction accuracy, suggesting that some metabolite levels are under non‐additive genetic control. Joint analysis of multiple metabolite accumulation simultaneously was effective in improving prediction accuracy by exploiting correlations among metabolites. The current study serves as an important first step in evaluating the cumulative effect of markers in influencing metabolic variation under control and HNT conditions.

show abstract

MegaLMM: Mega-scale linear mixed models for genomic predictions with thousands of traits

Cited by 47 publications

References 84 publications

Bayesian multitrait kernel methods improve multienvironment genome-based prediction

Bayesian multitrait kernel methods improve multienvironment genome-based prediction

Predicting Genotype × Environment × Management (G × E × M) Interactions for the Design of Crop Improvement Strategies

Genomic prediction of metabolic content in rice grain in response to warmer night conditions

Contact Info

Product

Resources

About