Performance evaluation of support vector machine (SVM)-based predictors in genomic selection

SEYED AMIR KASNAVI; MAHDI AMIN AFSHAR; MOHAMMAD MAHDI SHARIATI; NASSER EMAM JOMEH KASHAN; MAHMOOD HONARVAR

doi:10.56093/ijans.v87i10.75270

Authors

SEYED AMIR KASNAVI Islamic Azad University, Tehran, Iran
MAHDI AMIN AFSHAR Islamic Azad University, Tehran, Iran
MOHAMMAD MAHDI SHARIATI Islamic Azad University, Tehran, Iran
NASSER EMAM JOMEH KASHAN Islamic Azad University, Tehran, Iran
MAHMOOD HONARVAR Islamic Azad University, Tehran, Iran

https://doi.org/10.56093/ijans.v87i10.75270

Keywords:

Genetic architecture, Genomic breeding values, QTL effects, SNP, Support vector machine

Abstract

The aim was to compare predictive performance of SVM-based predictors constructed using different kernel functions (radial, sigmoid, linear and polynomial) in different genetic architectures of a trait (number of QTL, distribution of QTL effects) and heritability levels. To this end, a genome comprised of five chromosomes, one
Morgan each, was simulated on which 10,000 bi-allelic single nucleotide polymorphisms (SNP) were distributed.
Cross validation employing a grid search was used to tune the meta-parameters of each kernel function. Pearson’s
correlation between the true and predicted genomic breeding values (rp,t) and mean squared error of predicted
genomic breeding values (MSEp) were used, respectively, as measures of the predictive accuracy and the overall
fit. Meta-parameter optimization had a significant effect on predictive performance of SVM-based predictors in
such a way that by using improper meta-parameters, the predictive power of models decreased significantly. In all
models, the accuracy of prediction increased following increase in heritability and decrease in the number of
QTLs. In most of scenarios, radial- and sigmoid-based SVM predictors outperformed polynomial and linear models.
The linear-and polynomial-based SVM had lower rp,t and higher MSEp and, therefore, were not recommended for
genomic selection. The prediction accuracy of radial and sigmoid models was approximately the same in most of
the studied scenarios; however, considering all pros and cons of radial and sigmoid kernels, radial kernel was
recommended as the best kernel function for constructing SVM. All of studied SVM-based predictors were efficient
users of time and memory.

Downloads

Download data is not yet available.

References

Blondel M, Onogi A, Iwata H and Ueda N. 2015. A ranking approach to genomic selection.PLoS ONE 10(6): e0128570. DOI: https://doi.org/10.1371/journal.pone.0128570

Boser B, Guyon I and Vapnik V. 1992. An training algorithm for optimal margin classifiers. Proceedings of the Fifth Annual Workshop on Computational Learning Theory, pp 263–68. 27– 29 July 1992. Pittsburgh, USA. DOI: https://doi.org/10.1145/130385.130401

Combs E and Bernardo R. 2015. Accuracy of genome wide selection for different traits with constant population size, heritability, and number of markers. Plant Genome 6: 1. Daetwyler H D, Calus M P L, Pong-Wong R, de los Campos G and Hickey J M. 2013. Genomic prediction in animals and plants: simulation of data, validation, reporting, and benchmarking. Genetics 193: 347–65. DOI: https://doi.org/10.1534/genetics.112.147983

Ghafouri-Kesbi F, Rahimi-Mianji G, Honarvar M and Nejati- Javaremi A. 2016. Predictive ability of random forests, boosting, support vector machines and genomic best linear unbiased prediction in different scenarios of genomic evaluation. Animal Production Science 57: 229–36. DOI: https://doi.org/10.1071/AN15538

Hastie T J, Tibshirani R and Friedman J. 2009. The Elements of Statistical Learning. 745 p, Springer, New York, USA. DOI: https://doi.org/10.1007/978-0-387-84858-7

Hayes B J and Daetwyler H D. 2015. Genomic selection.Course note. February 2015. Armidale, Australia. Available at: http://jvanderw.une.edu.au/GenomicPredictionCoursNotes Armidale2015.pdf.

Howard R, Carriquiry A L and Beavis W D. 2014. Parametric and nonparametric statistical methods for genomic selection of traits with additive and epistatic genetic architectures. Genetics 4: 1027–46. DOI: https://doi.org/10.1534/g3.114.010298

Honarvar M and Ghiasi H. 2013. A comparison of genomic predictions using support vector machines (SVMs) and GBLUP methods. Agrochimica Research 57: 3–21.

Meuwissen T H E, Hayes B J and Goddard M E. 2001. Prediction of total genetic value using genome wide densemarker maps.Genetics 157: 1819–29. DOI: https://doi.org/10.1093/genetics/157.4.1819

Meyer D, Dimitriadou E, Hornik K, Weingessel A and Leisch K. 2013. Misc functions of the department of statistics (e1071), TU Wien. Available at: http://cran.r-project.org/web/packages/e1071/index.html.

Neves H H R, Carvalheiro R and Queiroz S A. 2012. A comparison of statistical methods for genomic selection in a mice population.BMC Genetics 13: 100. DOI: https://doi.org/10.1186/1471-2156-13-100

Scholkopf B, Tsuda K and Vert J P. 2004. Kernel methods in computational biology. MIT Press series on Computational Molecular Biology. 425 pp, MIT Press. Cambridge, Massachuse. DOI: https://doi.org/10.7551/mitpress/4057.001.0001

Technow F. 2013. hypred: Simulation of genomic data in applied genetics. Available at: http://cran.r-project.org/web/packages/hypred/index.html.

Zhu Y, Tan Y, Hua Y, Wang M, Zhang G and Zhang. 2010. Feature selection and performance evaluation of support vector machine (SVM)-based classifier for differentiating benign and malignant pulmonary nodules by computed tomography. Journal of Digital Imaging 23: 51–65. DOI: https://doi.org/10.1007/s10278-009-9185-9