image: Relative contribution of different model components estimated for eleven traits. A, Average proportions of phenotypic variance related to genotypic (g) and genomic (G) effects, their interactions (×) with the vector of environments (E), the enviromic effects (W), the interaction effects G × W, as well as the residual effect extracted from the statistical genomic prediction model fits. The relationship matrices for the different effects in the statistical genomic prediction models were constructed using the G-BLUP approach or, where indicated, the Gaussian kernel (GK) or Deep kernel (DK). The statistical genomic prediction models were compared with a model based on phenotypic data (Phenotypic). Error bars correspond to standard deviation around the mean. B, Average proportions of phenotypic variance related to genomic (G), additive (A), and dominance (D) effects, their interactions (×) with the vector of environments (E), and the residual effect extracted from the statistical genomic prediction model fits. The model structures G and G + D were additionally extended with the fixed effect of inbreeding (inb). The relationship matrices for the different effects were based on G-BLUP. Error bars correspond to standard deviation around the mean. The results for the benchmark model G are the same as shown in A. C Relative contribution of the SNP, PC, weather, and soil feature streams estimated using SHAP for the deep learning genomic prediction model. Error bars correspond to standard deviation around the mean.
Credit: Horticulture Research
In the face of climate change, apple breeding programs need innovative ways to select cultivars that can thrive under diverse conditions. A recent study demonstrates how integrating genomic data with environmental factors such as weather and soil can drastically improve the selection process. By using multi-environmental genomic prediction models, including deep learning, researchers have enhanced the accuracy of predictions for key apple traits. These advancements provide a promising tool for breeding apple varieties that are not only high-yielding but also adaptable to the fluctuating climate, offering a potential breakthrough in ensuring future food security.
Traditional apple breeding methods often fail to account for the complexities of genotype-by-environment interactions, which are crucial in selecting the best cultivars for varying climates. To overcome this limitation, the latest research combines phenotypic, genomic, and environmental data into sophisticated multi-environmental prediction models. These models are designed to predict how apple cultivars will perform across different environments, a process that has historically been difficult due to the vast diversity of environmental conditions. This new approach leverages both statistical and deep learning methods to process complex datasets, offering a more robust and accurate model for future breeding efforts.
Published (DOI: 10.1093/hr/uhae319) in Horticulture Research (November 2024), this study from Agroscope, ETH Zurich, and collaborators presents a breakthrough in apple breeding by applying multi-environmental genomic prediction. The research explores the integration of genomic data with environmental variables such as soil conditions and weather patterns, providing a new method for predicting key apple traits. By incorporating genotype-by-environment interactions into the prediction models, the study paves the way for selecting apple cultivars that are better suited to different climates, ultimately helping breeders respond to the challenges posed by climate change.
The study utilized the apple REFPOP, a comprehensive genetic population, to examine how different models predict 11 important traits in apples, including harvest date, fruit weight, and acidity. Researchers employed a range of prediction techniques, from the standard G-BLUP method to more advanced deep learning models, assessing the impact of genotype-by-environment interactions. The results showed that incorporating environmental factors such as weather and soil significantly improved the prediction accuracy for most traits, particularly when the G-BLUP model was enriched with environmental data.
The study also highlighted the power of deep learning models, which outperformed traditional methods for traits with complex genetic architectures, such as harvest date and titratable acidity. For these traits, deep learning models improved predictive ability by up to 0.10, providing a clear advantage in precision. The findings emphasize that as climate variability becomes more pronounced, integrating both genomic and environmental data using advanced machine learning approaches will be key to breeding apple cultivars that can adapt to future environmental challenges.
“By combining genomic data with environmental factors, we are opening a new frontier in apple breeding,” said Dr. Michaela Jung, the lead researcher from Agroscope. “The ability to predict how different apple cultivars will perform under various environmental conditions will give breeders a powerful tool to select varieties that are not only high-yielding but also climate-resilient. Deep learning models, in particular, have shown immense potential in refining these predictions, offering a promising solution for adapting apple breeding to the challenges of a changing climate.”
This study provides a roadmap for apple breeders to develop cultivars that are more resilient to climate change, ensuring stable production despite fluctuating environmental conditions. The integration of genomic and environmental data in breeding programs will enable the selection of apple varieties that are better suited for different regions and climates, improving both yield and quality. Moreover, the application of deep learning models in multi-environmental genomic prediction can be extended to other crops, offering a broader solution for global food security. By harnessing big data and advanced algorithms, this approach could revolutionize the way crops are bred, making them more adaptable to the future.
###
References
DOI
Original Source URL
https://doi.org/10.1093/hr/uhae319
Funding information
C.Q.-T. was supported by the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 847585 – RESPONSE. This study was partially funded by the FOAG project ‘Apfelzukunft dank Züchtung’ (2020/17/AZZ).
About Horticulture Research
Horticulture Research is an open access journal of Nanjing Agricultural University and ranked number one in the Horticulture category of the Journal Citation Reports ™ from Clarivate, 2023. The journal is committed to publishing original research articles, reviews, perspectives, comments, correspondence articles and letters to the editor related to all major horticultural plants and disciplines, including biotechnology, breeding, cellular and molecular biology, evolution, genetics, inter-species interactions, physiology, and the origination and domestication of crops.
Journal
Horticulture Research
Subject of Research
Not applicable
Article Title
Integrative multi-environmental genomic prediction in apple
Article Publication Date
20-Nov-2024
COI Statement
The authors declare that they have no competing interests.