Deep Kernel and Deep Learning for Genome-Based Prediction of Single Traits in Multienvironment Breeding Trials

Authors

José Crossa, International Maize and Wheat Improvement Center (CIMMYT) & Estadistica e Informatica, Colegio de Postgraduados
Johannes W.R. Martini, International Maize and Wheat Improvement Center (CIMMYT)
Daniel Gianola, University of Wisconsin- Madison
Paulino Pérez-Rodríguez, Colegio de Postgraduados
Diego Jarquin, University of Nebraska-LincolnFollow
Philomin Juliana, International Maize and Wheat Improvement Center (CIMMYT)
Osval Antonio Montesinos López, University of Nebraska – Lincoln & Universidad de ColimaFollow
Jaime Cuevas, Universidad de Quintana RooFollow

Date of this Version

12-9-2019

Citation

Crossa J, Martini JWR, Gianola D, Pérez-Rodríguez P, Jarquin D, Juliana P, Montesinos-López O and Cuevas J (2019) Deep Kernel and Deep Learning for Genome- Based Prediction of Single Traits in Multienvironment Breeding Trials. Front. Genet. 10:1168. doi: 10.3389/fgene.2019.01168

Abstract

Deep learning (DL) is a promising method for genomic-enabled prediction. However, the implementation of DL is difficult because many hyperparameters (number of hidden layers, number of neurons, learning rate, number of epochs, batch size, etc.) need to be tuned. For this reason, deep kernel methods, which only require defining the number of layers, may be an attractive alternative. Deep kernel methods emulate DL models with a large number of neurons, but are defined by relatively easily computed covariance matrices. In this research, we compared the genome-based prediction of DL to a deep kernel (arc-cosine kernel, AK), to the commonly used non-additive Gaussian kernel (GK), as well as to the conventional additive genomic best linear unbiased predictor (GBLUP/ GB). We used two real wheat data sets for benchmarking these methods. On average, AK and GK outperformed DL and GB. The gain in terms of prediction performance of AK and GK over DL and GB was not large, but AK and GK have the advantage that only one parameter, the number of layers (AK) or the bandwidth parameter (GK), has to be tuned in each method. Furthermore, although AK and GK had similar performance, deep kernel AK is easier to implement than GK, since the parameter “number of layers” is more easily determined than the bandwidth parameter of GK. Comparing AK and DL for the data set of year 2015–2016, the difference in performance of the two methods was bigger, with AK predicting much better than DL. On this data, the optimization of the hyperparameters for DL was difficult and the finally used parameters may have been suboptimal. Our results suggest that AK is a good alternative to DL with the advantage that practically no tuning process is required.