EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Critical evaluation of the theory and practice of feed-forward neural networks for genomic prediction

Aaron Kusmec, Karlene L Negus, Jianming Yu iD

G3: Genes, Genomes, Genetics · 04.03.2026 · 16 (3)

DOI: 10.1093/g3journal/jkaf314 ↗

Kurz gefasst

Fachartikel in der Zeitschrift G3: Genes, Genomes, Genetics, online erschienen am 24.12.2025. Verfasst von Aaron Kusmec und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Agronomy, Kansas State University, Manhattan, KS 66506-0110, United States. Hauptthemen laut MEDLINE: Genomics, Neural Networks, Computer. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 73 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Kansas State University, Agriculture and Food Research Initiative, USDA National Institute of Food and Agriculture.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Genomic prediction (GP) has catalyzed increased rates of genetic gain in animal and plant breeding. Recently, deep learning (DL) has been explored to increase GP accuracy by incorporating diverse data types and learning complex, non-linear patterns in datasets. However, DL consistently fails to significantly improve prediction accuracy over gold standard genomic BLUP (gBLUP) models. In this study, we first review the theory behind neural networks and reproducing kernel Hilbert spaces (RKHS) regression to contextualize 3 claimed benefits of DL over linear models: incorporation of diverse data types, avoidance of feature engineering, and universal approximation behavior. We also propose a taxonomy of prediction problems so that model comparisons do not confound differences in the predictive skill of different model classes with differences in the input data. Second, we leverage a maize multi-environment trial dataset to train DL and RKHS models that implicitly capture non-linear patterns between genomic, soil, weather, and management inputs and grain yield. The results demonstrate that feature engineering using principal components of SNPs generally degrades prediction accuracy across model classes. Furthermore, DL models persistently fail to outperform RKHS models across prediction problems. Finally, we evaluate the theoretical critiques with the empirical results, confirming the theoretical arguments. Nevertheless, a small portion of the possible DL model space has been explored, leaving open the possibility of DL making significant contributions to GP problems through additional aspects not considered here. We conclude by suggesting several avenues for further theoretical and practical research, including the resolution of several disciplinary differences.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

GenomicsNeural Networks, ComputerModels, GeneticZea maysDeep LearningPolymorphism, Single NucleotideFeedforward Neural Networks

Stichwörter der Autor:innen

deep learninggenomic predictionlinear modelsphenotypic prediction

Autor:innen und Institutionen

  • Department of Agronomy, Kansas State University, Manhattan, KS 66506-0110, United States.
  • Department of Agronomy, Iowa State University, Ames, IA 50011-1051, United States.

ORCID-Profile: Jianming Yu

Förderung

  • Kansas State University
  • Agriculture and Food Research Initiative
  • USDA National Institute of Food and Agriculture
  • Iowa State University Plant Sciences Institute
  • Beocat Research Cluster
  • NSF

Zitieren

APA

Kusmec, A., Negus, K. L., & Yu, J. (2026). Critical evaluation of the theory and practice of feed-forward neural networks for genomic prediction. G3: Genes, Genomes, Genetics, 16(3). https://doi.org/10.1093/g3journal/jkaf314

Vancouver

Kusmec A, Negus KL, Yu J. Critical evaluation of the theory and practice of feed-forward neural networks for genomic prediction. G3: Genes, Genomes, Genetics. 2026;16(3). doi:10.1093/g3journal/jkaf314

BibTeX
@article{Kusmec2026,
  author = {Kusmec, Aaron and Negus, Karlene L and Yu, Jianming},
  title = {Critical evaluation of the theory and practice of feed-forward neural networks for genomic prediction},
  journal = {G3: Genes, Genomes, Genetics},
  year = {2026},
  volume = {16},
  number = {3},
  doi = {10.1093/g3journal/jkaf314}
}
Kostenlos starten →