Impact Factor 3.517 | CiteScore 3.60
More on impact ›

Original Research ARTICLE Provisionally accepted The full-text will be published soon. Notify me

Front. Genet. | doi: 10.3389/fgene.2019.01168

DEEP KERNEL AND DEEP LEARNING FOR GENOME-BASED PREDICTION OF SINGLE TRAITS IN MULTI-ENVIRONMENT BREEDING TRIALS

 Jose Crossa1,  Johannes Martini2, Daniel Gianola3, Osval Montesinos-Lopez4*, Jaime Cuevas5*, Paulino Parez6, Diego jarquin7 and Philomin Juliana2
  • 1Biometrics and Statistics Unit, International Maize and Wheat Improvement Center (Mexico), Mexico
  • 2International Maize and Wheat Improvement Center (Mexico), Mexico
  • 3University of Wisconsin-Madison, United States
  • 4University of Colima, Mexico
  • 5University of Quintana Roo, Mexico
  • 6Graduate College (COLPOS), Mexico
  • 7University of Nebraska-Lincoln, United States

Deep learning (DL) is a promising method for genomic-enabled prediction. However, the implementation of DL is awkward because many hyper-parameters (number of hidden layers, number of neurons, learning rate, number of epochs, batch size, etc.) need to be tuned. On the other hand, deep kernel methods only require defining the number of layers for emulating DL models based on covariance matrices with a large number of neurons. In this research we compare 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. Our results suggest that AK is a good alternative to DL with the advantage that practically no tuning process is required.

Keywords: deep learning, Deep kernel, genomic-enabled predction, single trit, Multi-environment

Received: 16 Sep 2019; Accepted: 23 Oct 2019.

Copyright: © 2019 Crossa, Martini, Gianola, Montesinos-Lopez, Cuevas, Parez, jarquin and Juliana. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

* Correspondence:
Dr. Osval Montesinos-Lopez, University of Colima, Colima, 28040, Colima6, Mexico, o.montesinos@gmail.com
Dr. Jaime Cuevas, University of Quintana Roo, Chetumal, 77019, Quintana Roo, Mexico, j.cuevas@gamial.com