Deep Neural Networks for Genomic Prediction Do Not Estimate Marker Effects

Ubbens, Jordan; Parkin, Isobel A. P.; Eynck, Christina; Stavness, Ian; Sharpe, Andrew

doi:10.1101/2021.05.20.445038

Cited by 4 publications

(5 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead, it performs comparably to other non-linear methods, such as the non-linear SVM and the Decision Tree Classifier, even with the regularization features ( i.e ., batch normalization and dropout) that aid the same architecture in achieving top-5 success. These results are consistent with other work that show linear techniques outperform or match performance of more complex non-linear methods [5], [10], [11]. We believe that linear models outperform non-linear ones because the relationship between dog SNP sequences and breeds, given enough genomic positions, is additive.…”

Section: Resultssupporting

confidence: 91%

See 1 more Smart Citation

Predicting Dog Phenotypes from Genotypes

Bartusiak

Barrabes

Rymbekova

et al. 2022

Preprint

View full text Add to dashboard Cite

We analyze dog genotypes (i.e., positions of dog DNA sequences that often vary between different dogs) in order to predict the corresponding phenotypes (i.e., unique observed characteristics). More specifically, given chromosome data from a dog, we aim to predict the breed, height, and weight. We explore a variety of linear and non-linear classification and regression techniques to accomplish these three tasks. We also investigate the use of a neural network (both in linear and non-linear modes) for breed classification and compare the performance to traditional statistical methods. We show that linear methods generally outperform or match the performance of non-linear methods for breed classification. However, we show that the reverse is true for height and weight regression. Finally, we evaluate the results of all of these methods based on the number of input features used in the analysis. We conduct experiments using different fractions of the full genomic sequences, resulting in input sequences ranging from 20 SNPs to ~200k SNPs. In doing so, we explore the impact of using a very limited number of SNPs for prediction. Our experiments demonstrate that these phenotypes in dogs can be predicted with as few as 0.5% of randomly selected SNPs (i.e., 992 SNPs) and that dog breeds can be classified with 50% balanced accuracy with as few as 0.02% SNPs (i.e., 40 SNPs).

show abstract

Section: Resultssupporting

confidence: 91%

“…On the other hand, non-linear techniques show strong performance for regression tasks. Recent work confirms these trends [5], [10], [11].…”

Section: Introductionsupporting

confidence: 58%

Predicting Dog Phenotypes from Genotypes

Bartusiak

Barrabes

Rymbekova

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Another consideration is recent evidence to suggest that DL models do not estimate complex marker effects, but rather use the genetic relatedness between markers to make predictions. This may partially explain the underperformance of DL in crop genomic prediction problems [38].…”

Section: Machine Learning Performancementioning

confidence: 99%

Machine learning models outperform deep learning models, provide interpretation and facilitate feature selection for soybean trait prediction

Gill

Anderson

et al. 2022

BMC Plant Biol

View full text Add to dashboard Cite

Recent growth in crop genomic and trait data have opened opportunities for the application of novel approaches to accelerate crop improvement. Machine learning and deep learning are at the forefront of prediction-based data analysis. However, few approaches for genotype to phenotype prediction compare machine learning with deep learning and further interpret the models that support the predictions. This study uses genome wide molecular markers and traits across 1110 soybean individuals to develop accurate prediction models. For 13/14 sets of predictions, XGBoost or random forest outperformed deep learning models in prediction performance. Top ranked SNPs by F-score were identified from XGBoost, and with further investigation found overlap with significantly associated loci identified from GWAS and previous literature. Feature importance rankings were used to reduce marker input by up to 90%, and subsequent models maintained or improved their prediction performance. These findings support interpretable machine learning as an approach for genomic based prediction of traits in soybean and other crops.

show abstract

“…In a recent study 29 it has been proposed the principle of shortcut learning to explain why deep learning does not outperform penalized regression methods. Briefly, it is theorized that neural networks tend to base their predictions on overall genetic relatedness rather than on the effects of specific markers with, say, epistatic effects.…”

Section: Comparison To the State Of The Artmentioning

confidence: 99%

Stacked kinship CNN vs. GBLUP for genomic predictions of additive and complex continuous phenotypes

Biscarini

2022

Preprint

View full text Add to dashboard Cite

Deep learning is impacting many fields of data science with often spectacular results. However, its application to whole-genomepredictions in plant and animal science or in human biology has been rather limited, with mostly underwhelming results.While most works focus on exploring alternative network architectures, in this study we propose an innovative representation ofmarker genotype data. Different types of genomic kinship matrices are stacked together in a 3D pile from where 2D grayscaleslices are extracted and fed to a deep convolutional neural network (DNN). We tested nine scenarios of simulated phenotypeswith combinations of additivity, dominance and epistasis, and compared the DNN to GBLUP-A (Genomic BLUP computedusing only the additive kinship matrix) and GBLUP-optim (Genomic BLUP with additive, dominance, and epistasis kinshipmatrices, as needed).Results varied depending on the accuracy metric employed, with DNN performing better in terms of Root Mean Squared Error(1%-12% lower than GBLUP-A; 1%-9% lower than GBLUP-optim) but worse in terms of Pearson’s correlation (0.505 for DNNcompared to 0.672 and 0.669 of GBLUP-A and GBLUP-optim for fully additive case; 0.274 for DNN, 0.279 for GBLUP-A, and0.477 for GBLUP-optim for fully dominant case).

show abstract

Deep Neural Networks for Genomic Prediction Do Not Estimate Marker Effects

Cited by 4 publications

References 20 publications

Predicting Dog Phenotypes from Genotypes

Predicting Dog Phenotypes from Genotypes

Machine learning models outperform deep learning models, provide interpretation and facilitate feature selection for soybean trait prediction

Stacked kinship CNN vs. GBLUP for genomic predictions of additive and complex continuous phenotypes

Contact Info

Product

Resources

About