A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species
Genomic selection is an integral tool for breeders to accurately select plants directly from genotype data leading to faster and more resource-efficient breeding programs. Several prediction methods have been established in the last few years. These range from classical linear mixed models to comple...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2022-11-01
|
| Series: | Frontiers in Plant Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fpls.2022.932512/full |
| _version_ | 1797990012011151360 |
|---|---|
| author | Maura John Maura John Florian Haselbeck Florian Haselbeck Rupashree Dass Christoph Malisi Patrizia Ricca Christian Dreischer Sebastian J. Schultheiss Dominik G. Grimm Dominik G. Grimm Dominik G. Grimm |
| author_facet | Maura John Maura John Florian Haselbeck Florian Haselbeck Rupashree Dass Christoph Malisi Patrizia Ricca Christian Dreischer Sebastian J. Schultheiss Dominik G. Grimm Dominik G. Grimm Dominik G. Grimm |
| author_sort | Maura John |
| collection | DOAJ |
| description | Genomic selection is an integral tool for breeders to accurately select plants directly from genotype data leading to faster and more resource-efficient breeding programs. Several prediction methods have been established in the last few years. These range from classical linear mixed models to complex non-linear machine learning approaches, such as Support Vector Regression, and modern deep learning-based architectures. Many of these methods have been extensively evaluated on different crop species with varying outcomes. In this work, our aim is to systematically compare 12 different phenotype prediction models, including basic genomic selection methods to more advanced deep learning-based techniques. More importantly, we assess the performance of these models on simulated phenotype data as well as on real-world data from Arabidopsis thaliana and two breeding datasets from soy and corn. The synthetic phenotypic data allow us to analyze all prediction models and especially the selected markers under controlled and predefined settings. We show that Bayes B and linear regression models with sparsity constraints perform best under different simulation settings with respect to explained variance. Further, we can confirm results from other studies that there is no superiority of more complex neural network-based architectures for phenotype prediction compared to well-established methods. However, on real-world data, for which several prediction models yield comparable results with slight advantages for Elastic Net, this picture is less clear, suggesting that there is a lot of room for future research. |
| first_indexed | 2024-04-11T08:28:39Z |
| format | Article |
| id | doaj.art-1989c900845642afbd2043b4c21e4580 |
| institution | Directory Open Access Journal |
| issn | 1664-462X |
| language | English |
| last_indexed | 2024-04-11T08:28:39Z |
| publishDate | 2022-11-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Plant Science |
| spelling | doaj.art-1989c900845642afbd2043b4c21e45802022-12-22T04:34:36ZengFrontiers Media S.A.Frontiers in Plant Science1664-462X2022-11-011310.3389/fpls.2022.932512932512A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant speciesMaura John0Maura John1Florian Haselbeck2Florian Haselbeck3Rupashree Dass4Christoph Malisi5Patrizia Ricca6Christian Dreischer7Sebastian J. Schultheiss8Dominik G. Grimm9Dominik G. Grimm10Dominik G. Grimm11Technical University of Munich, Campus Straubing for Biotechnology and Sustainability, Bioinformatics, Straubing, GermanyWeihenstephan-Triesdorf University of Applied Sciences, Bioinformatics, Straubing, GermanyTechnical University of Munich, Campus Straubing for Biotechnology and Sustainability, Bioinformatics, Straubing, GermanyWeihenstephan-Triesdorf University of Applied Sciences, Bioinformatics, Straubing, GermanyComputomics GmbH, Tübingen, GermanyComputomics GmbH, Tübingen, GermanyComputomics GmbH, Tübingen, GermanyComputomics GmbH, Tübingen, GermanyComputomics GmbH, Tübingen, GermanyTechnical University of Munich, Campus Straubing for Biotechnology and Sustainability, Bioinformatics, Straubing, GermanyWeihenstephan-Triesdorf University of Applied Sciences, Bioinformatics, Straubing, GermanyTechnical University of Munich, Department of Informatics, Garching, GermanyGenomic selection is an integral tool for breeders to accurately select plants directly from genotype data leading to faster and more resource-efficient breeding programs. Several prediction methods have been established in the last few years. These range from classical linear mixed models to complex non-linear machine learning approaches, such as Support Vector Regression, and modern deep learning-based architectures. Many of these methods have been extensively evaluated on different crop species with varying outcomes. In this work, our aim is to systematically compare 12 different phenotype prediction models, including basic genomic selection methods to more advanced deep learning-based techniques. More importantly, we assess the performance of these models on simulated phenotype data as well as on real-world data from Arabidopsis thaliana and two breeding datasets from soy and corn. The synthetic phenotypic data allow us to analyze all prediction models and especially the selected markers under controlled and predefined settings. We show that Bayes B and linear regression models with sparsity constraints perform best under different simulation settings with respect to explained variance. Further, we can confirm results from other studies that there is no superiority of more complex neural network-based architectures for phenotype prediction compared to well-established methods. However, on real-world data, for which several prediction models yield comparable results with slight advantages for Elastic Net, this picture is less clear, suggesting that there is a lot of room for future research.https://www.frontiersin.org/articles/10.3389/fpls.2022.932512/fullphenotype predictiongenomic selectionplant phenotypingmachine learningArabidopsis thaliana |
| spellingShingle | Maura John Maura John Florian Haselbeck Florian Haselbeck Rupashree Dass Christoph Malisi Patrizia Ricca Christian Dreischer Sebastian J. Schultheiss Dominik G. Grimm Dominik G. Grimm Dominik G. Grimm A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species Frontiers in Plant Science phenotype prediction genomic selection plant phenotyping machine learning Arabidopsis thaliana |
| title | A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species |
| title_full | A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species |
| title_fullStr | A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species |
| title_full_unstemmed | A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species |
| title_short | A comparison of classical and machine learning-based phenotype prediction methods on simulated data and three plant species |
| title_sort | comparison of classical and machine learning based phenotype prediction methods on simulated data and three plant species |
| topic | phenotype prediction genomic selection plant phenotyping machine learning Arabidopsis thaliana |
| url | https://www.frontiersin.org/articles/10.3389/fpls.2022.932512/full |
| work_keys_str_mv | AT maurajohn acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT maurajohn acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT florianhaselbeck acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT florianhaselbeck acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT rupashreedass acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT christophmalisi acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT patriziaricca acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT christiandreischer acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT sebastianjschultheiss acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT dominikggrimm acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT dominikggrimm acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT dominikggrimm acomparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT maurajohn comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT maurajohn comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT florianhaselbeck comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT florianhaselbeck comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT rupashreedass comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT christophmalisi comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT patriziaricca comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT christiandreischer comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT sebastianjschultheiss comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT dominikggrimm comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT dominikggrimm comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies AT dominikggrimm comparisonofclassicalandmachinelearningbasedphenotypepredictionmethodsonsimulateddataandthreeplantspecies |