CoChat

Deep learning versus parametric and ensemble methods for genomic prediction of complex phenotypes

Open access
Abdollahi-Arpanahi, Rostam, Gianola, Daniel, Peñagaricano, Francisco · 2020 · Genetics Selection Evolution

The DOI matches its Crossref record; 51 of the 52 references it cites were found in a registry or an index.

Abstract

Background: Transforming large amounts of genomic data into valuable knowledge for predicting complex traits has been an important challenge for animal and plant breeders. Prediction of complex traits has not escaped the current excitement on machine-learning, including interest in deep learning algorithms such as multilayer perceptrons (MLP) and convolutional neural networks (CNN). The aim of this study was to compare the predictive performance of two deep learning methods (MLP and CNN), two ensemble learning methods [random forests (RF) and gradient boosting (GB)], and two parametric methods [genomic best linear unbiased prediction (GBLUP) and Bayes B] using real and simulated datasets. Methods: The real dataset consisted of 11,790 Holstein bulls with sire conception rate (SCR) records and genotyped for 58k single nucleotide polymorphisms (SNPs). To support the evaluation of deep learning methods, various simulation studies were conducted using the observed genotype data as template, assuming a heritability of 0.30 with either additive or non-additive gene effects, and two different numbers of quantitative trait nucleotides (100 and 1000). Results: In the bull dataset, the best predictive correlation was obtained with GB (0.36), followed by Bayes B (0.34), GBLUP (0.33), RF (0.32), CNN (0.29) and MLP (0.26). The same trend was observed when using mean squared error of prediction. The simulation indicated that when gene action was purely additive, parametric methods outperformed other methods. When the gene action was a combination of additive, dominance and of two-locus epistasis, the best predictive ability was obtained with gradient boosting, and the superiority of deep learning over the parametric methods depended on the number of loci controlling the trait and on sample size. In fact, with a large dataset including 80k individuals, the predictive performance of deep learning methods was similar or slightly better than that of parametric methods for traits with non-additive gene action. Conclusions: For prediction of traits with non-additive gene action, gradient boosting was a robust method. Deep learning approaches were not better for genomic prediction unless non-additive variance was sizable.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1186/s12711-020-00531-z

Cites

51 of 51 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 a Representation of a multilayer perceptron (MLP) network. Each unit is connected to the units of previous layers by a weighted linear summation, here represented by weight matrices Wi, and an activation function. Redrawn from: http://www.texample.net/tikz/examples/neural-network/. b Representation of a convolutional neural network (CNN). (i) The input layer consists of the SNP markers. (ii) Convolution layer consists of k filters, which capture the information in input layer by moving filters horizontally with a stride of “s” SNPs. (iii) Pooling layer involves of filters, combining the output of the previous convolution layer at certain locations into a single neuron. (iv) Fully connected layers connect every neuron in previous layer to every neuron in next layer. ‘ReLU’ indicates the rectified linear unit; softReLU indicates smooth rectified linear unit; Dropout indicates the dropout conduct layer
Fig. 1 a Representation of a multilayer perceptron (MLP) network. Each unit is connected to the units of previous layers by a weighted linear summation, here represented by weight matrices Wi, and an activation function. Redrawn from: http://www.texample.net/tikz/examples/neural-network/. b Representation of a convolutional neural network (CNN). (i) The input layer consists of the SNP markers. (ii) Convolution layer consists of k filters, which capture the information in input layer by moving filters horizontally with a stride of “s” SNPs. (iii) Pooling layer involves of filters, combining the output of the previous convolution layer at certain locations into a single neuron. (iv) Fully connected layers connect every neuron in previous layer to every neuron in next layer. ‘ReLU’ indicates the rectified linear unit; softReLU indicates smooth rectified linear unit; Dropout indicates the dropout conduct layerCC BY · PubMed Central
Fig. 2 Predictive correlation (left panel) and mean squared error of prediction (right panel) of two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using a real dataset of sire conception rate records from US Holstein bulls. The whiskers represent 95% confidence intervals
Fig. 2 Predictive correlation (left panel) and mean squared error of prediction (right panel) of two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using a real dataset of sire conception rate records from US Holstein bulls. The whiskers represent 95% confidence intervalsCC BY · PubMed Central
Fig. 3 Predictive ability of two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using genotypes at causal loci. Predictive ability was evaluated using predictive correlation a, b and mean squared error c, d. Different numbers of causal QTN (100 or 1000) and two scenarios of gene action, namely additive and a combination of additive, dominance and epistasis were investigated. The QTN were distributed as clustered across the entire genome.
Fig. 3 Predictive ability of two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using genotypes at causal loci. Predictive ability was evaluated using predictive correlation a, b and mean squared error c, d. Different numbers of causal QTN (100 or 1000) and two scenarios of gene action, namely additive and a combination of additive, dominance and epistasis were investigated. The QTN were distributed as clustered across the entire genome.CC BY · PubMed Central
Fig. 4 Predictive ability of two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using genotypes at marker loci. Predictive ability was evaluated using predictive correlation a, b and mean squared error c, d. Different numbers of QTN (100 or 1000) and two scenarios of gene action, namely additive and a combination of additive, dominance and epistasis were investigated. The QTN were distributed as clustered across the genome
Fig. 4 Predictive ability of two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using genotypes at marker loci. Predictive ability was evaluated using predictive correlation a, b and mean squared error c, d. Different numbers of QTN (100 or 1000) and two scenarios of gene action, namely additive and a combination of additive, dominance and epistasis were investigated. The QTN were distributed as clustered across the genomeCC BY · PubMed Central
Fig. 5 Predictive ability under two sample sizes, 12k and 80k individuals, for two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using genotypes at causal loci. Predictive ability was evaluated using predictive correlation a and mean squared error b. The 1000 causal QTN were distributed as clustered across the genome and gene action was a combination of additive, dominance and epistasis effects
Fig. 5 Predictive ability under two sample sizes, 12k and 80k individuals, for two conventional statistical methods (GBLUP and Bayes B) and four machine-learning methods including random forests (RF), gradient boosting (Boosting), multilayer perceptron (MLP) and convolutional neural network (CNN) using genotypes at causal loci. Predictive ability was evaluated using predictive correlation a and mean squared error b. The 1000 causal QTN were distributed as clustered across the genome and gene action was a combination of additive, dominance and epistasis effectsCC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 51 references

Cited by

See all 228

Metadata

Authors
Abdollahi-Arpanahi, R., Gianola, D., & Peñagaricano, F.
Affiliations
University of Florida (US)
Corresponding author
Peñagaricano, Francisco
Venue
Genetics Selection Evolution
Published
2020-02-24
Volume
52
Issue
1
Pages
12-12
ISSN
  • 0999-193X (ISSN-L)
  • 1297-9686
OpenAlex
W3012909962
Cited by
228
References
52
Citations by year
2026: 39 · 2025: 47 · 2024: 36 · 2023: 31 · 2022: 29 · 2021: 32 · 2020: 13
FWCI
15.39
Type
journal article
Language
en
Open access
Licence
Locations
  • Genetics Selection Evolution (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • HAL (submittedVersion, other-oa)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic and phenotypic traits in livestock
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Mapping and Diversity in Plants and Animals
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
Keywords
genomic prediction, complex traits, deep learning, multilayer perceptron, convolutional neural networks, random forest, gradient boosting, genomic best linear unbiased prediction, BayesB, Holstein bulls, non-additive gene action, dominance effects, epistasis, simulated data, prediction accuracy
MeSH terms
Deep Learning, Animals, Bayes Theorem, Cattle, Genotype, Models, Genetic, Phenotype, Quantitative Trait, Heritable, Multifactorial Inheritance, Polymorphism, Single Nucleotide, Genomics

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Deep learning versus parametric and ensemble methods for genomic prediction of complex phenotypes. https://app.cochat.ai/papers/10.1186/s12711-020-00531-z
The paper
Abdollahi-Arpanahi, R., Gianola, D., & Peñagaricano, F. (2020). Deep learning versus parametric and ensemble methods for genomic prediction of complex phenotypes. Genetics Selection Evolution. https://doi.org/10.1186/s12711-020-00531-z

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.