The optimal number of partial least squares components in genomic selection for pork pH

Silveira, Fernanda Gomes da; Duarte, Darlene Ana Souza; Chaves, Lucas Monteiro; Silva, Fabyano Fonseca e; Filho, Ivan Carvalho; Duarte, Márcio de Souza; Lopes, Paulo Sávio; Guimarães, Simone Eliza Facioni

doi:10.1590/0103-8478cr20151563

Cited by 3 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is important to highlight that the larger the data set, the more computational resources are required under the MT-PLS model because in order to select the optimal hyperparameters (number of principal components), an inner (nested) cross-validation needs to be implemented to select the optimal number of principal components. However, Silveira et al (2017) conducted research for selecting the optimal number of principal components in the context of UT-PLS using nested cross-validation as we did and using the degree-of-freedom (DoF) method and they reported that both approaches found the same optimal number of components. This option for selecting the optimal number of principal components should be explored in the context of MT-PLS since significant computational resources can be saved for implementing the MT-PLS method.…”

Section: Discussionmentioning

confidence: 99%

Multi-trait genome prediction of new environments with partial least squares

Montesinos‐López

Sandoval

et al. 2022

Front. Genet.

View full text Add to dashboard Cite

The genomic selection (GS) methodology proposed over 20 years ago by Meuwissen et al. (Genetics, 2001) has revolutionized plant breeding. A predictive methodology that trains statistical machine learning algorithms with phenotypic and genotypic data of a reference population and makes predictions for genotyped candidate lines, GS saves significant resources in the selection of candidate individuals. However, its practical implementation is still challenging when the plant breeder is interested in the prediction of future seasons or new locations and/or environments, which is called the “leave one environment out” issue. Furthermore, because the distributions of the training and testing set do not match, most statistical machine learning methods struggle to produce moderate or reasonable prediction accuracies. For this reason, the main objective of this study was to explore the use of the multi-trait partial least square (MT-PLS) regression methodology for this specific task, benchmarking its performance with the Bayesian Multi-trait Genomic Best Linear Unbiased Predictor (MT-GBLUP) method. The benchmarking process was performed with five actual data sets. We found that in all data sets the MT-PLS method outperformed the popular MT-GBLUP method by 349.8% (under predictor E + G), 484.4% (under predictor E + G + GE; where E denotes environments, G genotypes and GE the genotype by environment interaction) and 15.9% (under predictor G + GE) across traits. Our results provide empirical evidence of the power of the MT-PLS methodology for the prediction of future seasons or new environments. Furthermore, the comparison between single univariate-trait (UT) versus MT for GBLUP and PLS gave an increase in prediction accuracy of MT-GBLUP versus UT-GBLUP, but not for MT-PLS versus UT-PLS.

show abstract

Section: Discussionmentioning

confidence: 99%

Multi-trait genome prediction of new environments with partial least squares

Montesinos‐López

Sandoval

et al. 2022

Front. Genet.

View full text Add to dashboard Cite

show abstract

“…Furthermore, the applications in biometrics, environmental and agricultural data of regression methods based on dimensionality reduction are observed in the genomic selection (Azevedo et al, 2013a;Azevedo et al, 2013b;Azevedo et al, 2014;Azevedo et al, 2015;Silveira et al, 2017), gene expression (Nascimento et al, 2017), spectroscopy analysis NIR (Near Infrared) (Morgano et al, 2005;Teófilo et al, 2009), sensory data (Westad, 2005), climatic data (Lim et al, 2015) and process control data (Han et al, 2003), among others. However, dimensionality reduction methods can be applied whenever multicollinearity and high dimensionality are present, and the the study is aimed at regression.…”

Section: Introductionmentioning

confidence: 99%

Research Article A comparison of regression methods based on dimensional reduction for genomic prediction

Costa¹,

Azevedo²,

Nascimento³

et al. 2021

Genet. Mol. Res.

View full text Add to dashboard Cite

The quality of fit of a multiple linear regression model often encounters multicollinearity and high dimensionality problems, making it impossible to obtain stable estimates through the traditional method of estimation based on ordinary least squares. To overcome such challenges, dimensionality reduction methods have been proposed, because of their simple theory and easy application. We compared three dimensionality reduction methods: Principal Components Regression (PCR), Partial Least Squares (PLS), and Independent Components Regression (ICR). An important step for dimensionality reduction and prediction is selecting the number of components, as it affects the linear combinations of the explanatory variables. The linear combinations are inserted into the model to predict the response based on a reduced number of parameters. We examined the criteria for the selection of the number of components. The dimensionality reduction methods were applied to genomic and phenotype data. We evaluated 370 accessions of Asian rice, Oryza sativa, which were genotyped for 36,901 SNPs markers considered to predict the genomic values for the number of panicles per plant trait. ©FUNPEC-RP www.funpecrp.com.br Genetics and Molecular Research 20 (2): gmr18877 J.A. da Costa et al. 2This data set presented multicollinearity and high dimensionality.The computational time for each method was also recorded. Among the methods, PCR and ICR gave the highest accuracy values, with ICR standing out for presenting estimates of the least biased genomic values. However, ICR required more computational time than the other methodologies.

show abstract

Metabolomic spectra for phenotypic prediction of malting quality in spring barley

Guo

Jahoor

Jensen

et al. 2022

Sci Rep

View full text Add to dashboard Cite

We investigated prediction of malting quality (MQ) phenotypes in different locations using metabolomic spectra, and compared the prediction ability of different models, and training population (TP) sizes. Data of five MQ traits was measured on 2667 individual plots of 564 malting spring barley lines from three years and two locations. A total of 24,018 metabolomic features (MFs) were measured on each wort sample. Two statistical models were used, a metabolomic best linear unbiased prediction (MBLUP) and a partial least squares regression (PLSR). Predictive ability within location and across locations were compared using cross-validation methods. For all traits, more than 90% of the total variance in MQ traits could be explained by MFs. The prediction accuracy increased with increasing TP size and stabilized when the TP size reached 1000. The optimal number of components considered in the PLSR models was 20. The accuracy using leave-one-line-out cross-validation ranged from 0.722 to 0.865 and using leave-one-location-out cross-validation from 0.517 to 0.817. In conclusion, the prediction accuracy of metabolomic prediction of MQ traits using MFs was high and MBLUP is better than PLSR if the training population is larger than 100. The results have significant implications for practical barley breeding for malting quality.

show abstract

The optimal number of partial least squares components in genomic selection for pork pH

Abstract: Número ótimo de componentes nos quadrados mínimos parciais aplicados à seleção genômica para pH da carne suína

Cited by 3 publications

References 9 publications

Multi-trait genome prediction of new environments with partial least squares

Multi-trait genome prediction of new environments with partial least squares

Research Article A comparison of regression methods based on dimensional reduction for genomic prediction

Metabolomic spectra for phenotypic prediction of malting quality in spring barley

Contact Info

Product

Resources

About