Integrative multi-environmental genomic prediction in apple

Abstract

Genomic prediction for multiple environments can aid the selection of genotypes suited to specific soil and climate conditions. Methodological advances allow effective integration of phenotypic, genomic (additive, nonadditive), and large-scale environmental (enviromic) data into multi-environmental genomic prediction models. These models can also account for genotype-by-environment interaction, utilize alternative relationship matrices (kernels), or substitute statistical approaches with deep learning. However, the application of multi-environmental genomic prediction in apple remained limited, likely due to the challenge of building multi-environmental datasets and structurally complex models. Here, we applied efficient statistical and deep learning models for multi-environmental genomic prediction of eleven apple traits with contrasting genetic architectures by integrating genomic- and enviromic-based model components. Incorporating genotype-by-environment interaction effects into statistical models improved predictive ability by up to 0.08 for nine traits compared to the benchmark model. This outcome, based on Gaussian and Deep kernels, shows these alternatives can effectively substitute the standard genomic best linear unbiased predictor (G-BLUP). Including nonadditive and enviromic-based effects resulted in a predictive ability very similar to the benchmark model. The deep learning approach achieved the highest predictive ability for three traits with oligogenic genetic architectures, outperforming the benchmark by up to 0.10. Our results demonstrate that the tested statistical models capture genotype-by-environment interactions particularly well, and the deep learning models efficiently integrate data from diverse sources. This study will foster the adoption of multi-environmental genomic prediction to select apple cultivars adapted to diverse environmental conditions, providing an opportunity to address climate change impacts.

Acknowledgements

The authors thank the field technicians and staff at INRAe IRHS and Experimental Unit (UE Horti), Angers, France, the Fruit Breeding Group at Agroscope in Waedenswil, Switzerland, and technical staff at all apple REFPOP sites for the maintenance of the orchards and phenotypic data collection. Phenotypic data collection was partially supported by the Horizon 2020 Framework Program of the European Union under grant agreement No 817970 (project INVITE: ‘Innovations in plant variety testing in Europe to foster the introduction of new varieties better adapted to varying biotic and abiotic conditions and to more sustainable crop management practices’). 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).

Author contributions

This research was conceived by M.J., M.Roth, M.J.A., W.G., F.L., H.M., B.S., and A.P. M.J., M.R., E.H., N.P., L.L., and F.D. contributed to data collection. M.J. carried out the statistical data analysis with the support of M.Roth., G.B., and A.P. C.Q.-T. performed the deep learning analysis in consultation with M.J., S.Y., and B.S. M.J. and C.Q.-T. wrote the article in consultation with M.Roth, M.J.A., H.M., M.R., W.G., F.L., S.Y., B.S., G.B., and A.P. All authors have read the manuscript and approved the version to be published.

Data availability

All SNP genotypic data used in this study have been deposited in the INRAe dataset archive at https://doi.org/10.15454/IOPGYF and https://doi.org/10.15454/1ERHGX. The raw phenotypic data are available in the INRAe dataset archive at https://doi.org/10.15454/VARJYJ. The code underlying this article is available in GitHub at https://github.com/MichaelaJung/Integrative-prediction. The phenotypic, enviromic, and imputed genomic data formatted as input files for the provided code are available in Zenodo at 10.5281/zenodo.14191209.

Conflict of interests

The authors declare no conflicts of interest.

Supplementary Data

Supplementary data is available at Horticulture Research online.

References

[1] Meuwissen THE, Hayes BJ, Goddard ME. Prediction of total genetic value using genome-wide dense marker maps. Genetics. 2001;157:1819-29

[2] García-Ruiz A, Cole JB, VanRaden PM. et al.Changes in genetic selection differentials and generation intervals in US Holstein dairy cattle as a result of genomic selection. Proc Natl Acad Sci. 2016;113:E3995-4004

[3] Voss-Fels KP, Cooper M, Hayes BJ. Accelerating crop genetic gains with genomic selection. Theor Appl Genet. 2019;132:669-86

[4] VanRaden PM. Efficient methods to compute genomic predic-tions. J Dairy Sci. 2008;91:4414-23

[5] Jarquín D, Crossa J, Lacaze X. et al. A reaction norm model for genomic selection using high-dimensional genomic and envi-ronmental data. Theor Appl Genet. 2014;127:595-607

[6] Lopez-Cruz M, Crossa J, Bonnett D. et al. Increased prediction accuracy in wheat breeding trials using a marker × environ-ment interaction genomic selection model. G3 (Bethesda). 2015;5:569-82

[7] Kostick SA, Bernardo R, Luby JJ. Genomewide selection for fruit quality traits in apple: breeding insights gained from prediction and postdiction. Hortic Res. 2023;10:uhad088

[8] Kumar S, Chagné D, Bink MCAM. et al. Genomic selection for fruit quality traits in apple ( Malus × domestica Borkh.). PLoS One. 2012;7:e36674

[9] Migicovsky Z, Gardner KM, Money D. et al. Genome to phenome mapping in apple using historical data. Plant Genome. 2016;9:

[10] Muranty H, Troggio M, Sadok IB. et al. Accuracy and responses of genomic selection on key traits in apple breeding. Hortic Res. 2015;2:15060

[11] Jung M, Keller B, Roth M. et al. Genetic architecture and genomic predictive ability of apple quantitative traits across environ-ments. Hortic Res. 2022;9:uhac028

[12] Jung M, Roth M, Aranzana MJ. et al. The apple REFPOP—a refer-ence population for genomics-assisted breeding in apple. Hortic Res. 2020;7:189

[13] Granato I, Cuevas J, Luna-Vázquez F. et al. BGGE: a new package for genomic-enabled prediction incorporating genotype × envi-ronment interaction models. G3 (Bethesda). 2018;8:3039-47

[14] Pérez P, de los Campos G. Genome-wide regression and pre-diction with the BGLR statistical package. Genetics. 2014;198:483-95

[15] Cuevas J, Montesinos-López O, Juliana P. et al. Deep kernel for genomic and near infrared predictions in multi-environment breeding trials. G3 (Bethesda). 2019;9:2913-24

[16] Álvarez-Castro JM, Carlborg Ö. A unified model for functional and statistical epistasis and its application in quantitative trait loci analysis. Genetics. 2007;176:1151-67

[17] Vitezica ZG, Legarra A, Toro MA. et al. Orthogonal estimates of variances for additive, dominance, and epistatic effects in populations. Genetics. 2017;206:1297-307

[18] Vitezica ZG, Varona L, Legarra A. On the additive and domi-nant variance and covariance of individuals within the genomic selection scope. Genetics. 2013;195:1223-30

[19] Kumar S, Molloy C, Muñoz P. et al. Genome-enabled estimates of additive and nonadditive genetic variances and prediction of apple phenotypes across environments. G3 (Bethesda). 2015;5:2711-8

[20] Roth M, Beugnot A, Mary-Huard T. et al. Improving genomic pre-dictions with inbreeding and nonadditive effects in two admixed maize hybrid populations in single and multienvironment con-texts. Genetics. 2022;220:iyac018

[21] Yadav S, Wei X, Joyce P. et al. Improved genomic prediction of clonal performance in sugarcane by exploiting non-additive genetic effects. Theor Appl Genet. 2021;134:2235-52

[22] Cooper M, Messina CD, Podlich D. et al. Predicting the future of plant breeding: complementing empirical evaluation with genetic prediction. Crop Pasture Sci. 2014;65:311-36

[23] Resende RT, Piepho H-P, Rosa GJM. et al. Enviromics in breeding: applications and perspectives on envirotypic-assisted selection. Theor Appl Genet. 2021;134:95-112

[24] Montesinos-López OA, Montesinos-López A, Pérez-Rodríguez P. et al. A review of deep learning applications for genomic selec-tion. BMC Genomics. 2021;22:19

[25] Kick DR, Wallace JG, Schnable JC. et al. Yield prediction through integration of genetic, environment, and management data through deep learning. G3 (Bethesda). 2023;13:jkad006

[26] Jurado-Ruiz F, Rousseau D, Botía JA. et al. GenoDrawing: an autoencoder framework for image prediction from SNP markers. Plant Phenomics. 2023;5:0113

[27] Amadeu RR, Ferrão LFV, Oliveira I. et al. Impact of dominance effects on autotetraploid genomic prediction. Crop Sci. 2020;60:656-65

[28] Vitezica ZG, Reverter A, Herring W. et al. Dominance and epistatic genetic variances for litter size in pigs using genomic models. Genet Sel Evol. 2018;50:71

[29] Strandén I, Christensen OF. Allele coding in genomic evaluation. Genet Sel Evol. 2011;43:25

[30] Bianco L, Cestaro A, Sargent DJ. et al. Development and vali-dation of a 20K single nucleotide polymorphism (SNP) whole genome genotyping array for apple ( Malus × domestica Borkh). PLoS One. 2014;9:e110377

[31] Watts S, Migicovsky Z, Myles S.Large-scale apple GWAS reveals NAC18.1 as a master regulator of ripening traits. Fruit Res. 2023;3:32

[32] Takos AM, Jaffé FW, Jacob SR. et al. Light-induced expression of a MYB gene regulates anthocyanin biosynthesis in red apples. Plant Physiol. 2006;142:1216-32

[33] Verma S, Evans K, Guan Y. et al.Two large-effect QTLs, Ma and Ma3, determine genetic potential for acidity in apple fruit: breeding insights from a multi-family study. Tree Genet Genomes. 2019;15:18

[34] Rymenants M, van de Weg E, Auwerkerken A. et al. Detection of QTL for apple fruit acidity and sweetness using sensorial evaluation in multiple pedigreed full-sib families. Tree Genet Genomes. 2020;16:71

[35] Minamikawa MF, Kunihisa M, Moriya S. et al. Genomic prediction and genome-wide association study using combined genotypic data from different genotyping systems: application to apple fruit quality traits. Hortic Res. 2024;11:uhae131

[36] Bianco L, Cestaro A, Linsmith G. et al. Development and vali-dation of the Axiom®Apple480K SNP genotyping array. Plant J. 2016;86:62-74

[37] Browning SR, Browning BL. Rapid and accurate haplotype phas-ing and missing-data inference for whole-genome association studies by use of localized haplotype clustering. Am J Hum Genet. 2007;81:1084-97

[38] Muranty H, Denancé C, Feugey L. et al. Using whole-genome SNP data to reconstruct a large multi-generation pedigree in apple germplasm. BMC Plant Biol. 2020;20:2

[39] Daccord N, Celton J-M, Linsmith G. et al. High-quality de novo assembly of the apple genome and methylome dynamics of early fruit development. Nat Genet. 2017;49:1099-106

[40] Bates D, Mächler M, Bolker B. et al. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1-48

[41] Rodríguez-Álvarez MX, Boer MP, van Eeuwijk FA. et al. Correcting for spatial heterogeneity in plant breeding experiments with P- splines. Spat Stat. 2018;23:52-71

[42] Bernal-Vasquez A-M, Utz HF, Piepho H-P. Outlier detection methods for generalized lattices: a case study on the transition from ANOVA to REML. Theor Appl Genet. 2016;129:787-804

[43] Wei T, Simko V. R package “corrplot”: visualization of a correla-tion matrix. 2021.

[44] González-Camacho JM, de los Campos G, Pérez P. et al. Genome-enabled prediction of genetic values using radial basis function neural networks. Theor Appl Genet. 2012;125:759-71

[45] Lundberg SM, Lee S-I. A unified approach to interpreting model predictions. In: GuyonI, LuxburgUV, BengioS, WallachH, FergusR, VishwanathanS, GarnettR, eds. Advances in Neural Information Processing Systems (Vol. 30). Curran Associates, Inc., 2017,

[46] RCoreTeam. (2022). R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing.

[47] Costa-NetoG, Fritsche-NetoR, CrossaJ. Nonlinear kernels, dominance, and envirotyping data increase the accuracy of genome-based prediction in multi-environment trials. Heredity. 2021a;126:92-106

[48] GarnettR, GarnettR, GarnettR, et al. EnvRtype: a software to interplay enviromics and quantitative genomics in agriculture. G3 (Bethesda). 2021b;11:jkab040