{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,25]],"date-time":"2026-07-25T16:45:36Z","timestamp":1784997936782,"version":"3.55.0"},"reference-count":65,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T00:00:00Z","timestamp":1772064000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>Accurate in-season prediction of seed yield and seed composition traits such as oil and protein are useful for gaining accuracy and efficiency in soybean breeding. These predictions can also inform farmers, enabling them to improve their field management practices, and guide their market decisions. We report a Transformer-based deep learning framework built on 30 years of multi-environment performance data from the Northern and Southern Uniform Soybean Tests (UST) across North America. Unlike earlier studies on seed yield, oil and protein prediction that focus on limited years, regions, single modalities, we utilized a comprehensive dataset that includes weather, genotype, and management factors, ensuring a more holistic approach to soybean yield, oil, and protein prediction. Our model integrates multivariate time-series weather data with genotypic relationship information, maturity group, and geographic location, to predict variety performance in diverse environments. Our model captures complex temporal patterns associated with trait variability; showing high predictive accuracy (R2) of 77.6 \u00b1 0.2%, 63.9 \u00b1 4.7%, and 79.3 \u00b1 2.3% for seed yield, oil, and protein, respectively. Additionally, for seed yield, we also evaluated multiple interpretability methods to assess feature importance for predictor variables and critical growing timepoints, and solar radiation and temperature were noted as the key predictors. Overall, these results demonstrate the usefulness of a Transformer-based model in trait predictions, and the utility of large cooperative datasets from breeding programs.<\/jats:p>","DOI":"10.3389\/frai.2026.1750108","type":"journal-article","created":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T07:37:43Z","timestamp":1772091463000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":1,"title":["Transformer model to determine spatio-temporal relationships of variables, and interpretability for soybean seed yield, oil, and protein prediction"],"prefix":"10.3389","volume":"9","author":[{"given":"Timilehin T.","family":"Ayanlade","sequence":"first","affiliation":[{"name":"Department of Mechanical Engineering, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liza","family":"Van der Laan","sequence":"additional","affiliation":[{"name":"Department of Agronomy, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qisai","family":"Liu","sequence":"additional","affiliation":[{"name":"Department of Mechanical Engineering, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tryambak","family":"Gangopadhyay","sequence":"additional","affiliation":[{"name":"Department of Mechanical Engineering, Iowa State University","place":["Ames, IA, United States"]},{"name":"Amazon Web Services, Amazon","place":["Santa Clara, CA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Johnathon","family":"Shook","sequence":"additional","affiliation":[{"name":"Department of Agronomy, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Arti","family":"Singh","sequence":"additional","affiliation":[{"name":"Department of Agronomy, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Baskar","family":"Ganapathysubramanian","sequence":"additional","affiliation":[{"name":"Department of Mechanical Engineering, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Soumik","family":"Sarkar","sequence":"additional","affiliation":[{"name":"Department of Mechanical Engineering, Iowa State University","place":["Ames, IA, United States"]},{"name":"Department of Computer Science, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Asheesh K.","family":"Singh","sequence":"additional","affiliation":[{"name":"Department of Agronomy, Iowa State University","place":["Ames, IA, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,2,26]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"26446","DOI":"10.1038\/s41598-025-11780-7","article-title":"Predicting wheat yield using deep learning and multi-source environmental data","volume":"15","author":"Ashfaq","year":"2025","journal-title":"Sci. Rep"},{"key":"B2","doi-asserted-by":"publisher","first-page":"1408047","DOI":"10.3389\/fpls.2024.1408047","article-title":"Integrating multi-modal remote sensing, deep learning, and attention mechanisms for yield prediction in plant breeding experiments","volume":"15","author":"Aviles Toledo","year":"2024","journal-title":"Front. Plant Sci"},{"key":"B3","doi-asserted-by":"publisher","first-page":"299","DOI":"10.1007\/978-981-97-7762-4_13","article-title":"Multi-modal ai for ultra-precision agriculture","volume":"161","author":"Ayanlade","year":"2024","journal-title":"Stud. Big Data"},{"key":"B4","article-title":"Learning accurate and interpretable tree-based models","author":"Balcan","year":"2024","journal-title":"arXiv preprint arXiv:2405.15911"},{"key":"B5","doi-asserted-by":"publisher","first-page":"3665","DOI":"10.1002\/jsfa.11713","article-title":"Machine learning algorithms for soybean yield forecasting in the Brazilian Cerrado","volume":"102","author":"Barbosa Dos Santos","year":"2022","journal-title":"J. Sci. Food Agric"},{"key":"B6","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.eja.2016.03.002","article-title":"Contribution of incident solar radiation on leaves and pods to soybean seed weight and composition","volume":"77","author":"Bianculli","year":"2016","journal-title":"Eur. J. Agron"},{"key":"B7","doi-asserted-by":"publisher","first-page":"1","DOI":"10.5772\/17596","article-title":"Soybean yield formation: what controls it and how it can be improved","volume":"2011","author":"Board","year":"2011","journal-title":"Soybean Physiol. Biochem"},{"key":"B8","doi-asserted-by":"publisher","first-page":"213","DOI":"10.2134\/agronj1996.00021962008800050005x","article-title":"Potential uses and limitations of crop models","volume":"52","author":"Boote","year":"1996","journal-title":"Agric. Syst"},{"key":"B9","doi-asserted-by":"publisher","first-page":"3135","DOI":"10.1002\/csc2.21336","article-title":"Leveraging soil mapping and machine learning to improve spatial adjustments in plant breeding trials","volume":"61","author":"Carroll","year":"2024","journal-title":"Crop Sci"},{"key":"B10","doi-asserted-by":"publisher","first-page":"1731","DOI":"10.2134\/agronj2010.0219","article-title":"Soybean development and yield are influenced by planting date and environmental conditions in the southeastern coastal plain, United States","volume":"102","author":"Chen","year":"2010","journal-title":"Agron. J"},{"key":"B11","doi-asserted-by":"publisher","first-page":"e70142","DOI":"10.1002\/csc2.70142","article-title":"Comprehensive assessment of soybean seed composition from field trials spanning 22 us states and 24 years: Predictive insights","volume":"65","author":"Chiozza","year":"2025","journal-title":"Crop Sci"},{"key":"B12","doi-asserted-by":"publisher","first-page":"diaa016","DOI":"10.1093\/insilicoplants\/diaa016","article-title":"Modelling selection response in plant-breeding programs using crop models as mechanistic gene-to-phenotype (CGM-G2P) multi-trait link functions","volume":"3","author":"Cooper","year":"2020","journal-title":"In silico Plants"},{"key":"B13","doi-asserted-by":"publisher","first-page":"756","DOI":"10.1016\/j.tplants.2024.12.009","article-title":"Expanding genomic prediction in plant breeding: harnessing big data, machine learning, and advanced software","volume":"30","author":"Crossa","year":"2025","journal-title":"Trends Plant Sci"},{"key":"B14","article-title":"Eraser: a benchmark to evaluate rationalized NLP models","author":"DeYoung","year":"2019","journal-title":"arXiv preprint arXiv:1911.03429"},{"key":"B15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3561048","article-title":"Explainable AI (xAI): core ideas, techniques, and solutions","volume":"55","author":"Dwivedi","year":"2023","journal-title":"ACM Comput. Surv"},{"key":"B16","doi-asserted-by":"publisher","first-page":"1329","DOI":"10.2135\/cropsci2004.0029","article-title":"Air temperature during seed filling and soybean seed germination and vigor","volume":"45","author":"Egli","year":"2005","journal-title":"Crop Sci"},{"key":"B17","doi-asserted-by":"publisher","first-page":"100397","DOI":"10.1016\/j.rsase.2020.100397","article-title":"Estimation of soybean yield from machine learning techniques and multispectral RPAS imagery","volume":"20","author":"Eugenio","year":"2020","journal-title":"Remote Sens. Appl. Soc. Environ"},{"key":"B18","doi-asserted-by":"publisher","first-page":"e20519","DOI":"10.1002\/tpg2.20519","article-title":"Introducing CHiDO-a no code genomic prediction software implementation for the characterization and integration of driven omics","volume":"18","author":"Gonz\u00e1lez","year":"2024","journal-title":"Plant Genome"},{"key":"B19","doi-asserted-by":"publisher","first-page":"1371","DOI":"10.5194\/esd-12-1371-2021","article-title":"Impacts of compound hot-dry extremes on us soybean yields","volume":"12","author":"Hamed","year":"2021","journal-title":"Earth Syst. Dyn"},{"key":"B20","doi-asserted-by":"publisher","first-page":"108096","DOI":"10.1016\/j.compag.2023.108096","article-title":"On-farm soybean seed protein and oil prediction using satellite data","volume":"212","author":"Hernandez","year":"2023","journal-title":"Comput. Electron. Agric"},{"key":"B21","doi-asserted-by":"publisher","first-page":"1735","DOI":"10.1162\/neco.1997.9.8.1735","article-title":"Long short-term memory","volume":"9","author":"Hochreiter","year":"1997","journal-title":"Neural Comput"},{"key":"B22","doi-asserted-by":"publisher","first-page":"500","DOI":"10.2134\/agronj1985.00021962007700030031x","article-title":"Soyphen: Soybean growth stages modeled from temperature, daylength, and water availability1","volume":"77","author":"Hodges","year":"1985","journal-title":"Agron. J"},{"key":"B23","article-title":"Interpreting time series transformer models and sensitivity analysis of population age groups to covid-19 infections","author":"Islam","year":"2024","journal-title":"arXiv preprint arXiv:2401.15119"},{"key":"B24","doi-asserted-by":"publisher","first-page":"73","DOI":"10.1016\/S0167-8809(01)00358-9","article-title":"Adaptation and evaluation of the CROPGRO-soybean model to predict regional yield and production","volume":"93","author":"Jagtap","year":"2002","journal-title":"Agric. Ecosyst. Environ"},{"key":"B25","doi-asserted-by":"publisher","first-page":"1790","DOI":"10.2135\/cropsci2004.0602","article-title":"Dry matter accumulation predictors for optimal yield in soybean","volume":"45","author":"James","year":"2005","journal-title":"Crop Sci"},{"key":"B26","doi-asserted-by":"publisher","first-page":"166","DOI":"10.1093\/bfgp\/elq001","article-title":"Genomic selection in plant breeding: from theory to practice","volume":"9","author":"Jannink","year":"2010","journal-title":"Brief. Funct. Genomics"},{"key":"B27","doi-asserted-by":"publisher","first-page":"235","DOI":"10.1016\/S1161-0301(02)00107-7","article-title":"The dssat cropping system model","volume":"18","author":"Jones","year":"2003","journal-title":"Eur. J. Agron"},{"key":"B28","doi-asserted-by":"publisher","first-page":"621","DOI":"10.3389\/fpls.2019.00621","article-title":"Crop yield prediction using deep neural networks","volume":"10","author":"Khaki","year":"2019","journal-title":"Front. Plant Sci"},{"key":"B29","doi-asserted-by":"publisher","first-page":"252","DOI":"10.1007\/s00122-023-04470-3","article-title":"Models to estimate genetic gain of soybean seed yield from annual multi-environment field trials","volume":"136","author":"Krause","year":"2023","journal-title":"Theoret. Appl. Genet"},{"key":"B30","doi-asserted-by":"publisher","first-page":"497","DOI":"10.1038\/nclimate1832","article-title":"The critical role of extreme heat for maize production in the United States","volume":"3","author":"Lobell","year":"2013","journal-title":"Nat. Clim. Chang"},{"key":"B31","article-title":"\u201cA unified approach to interpreting model predictions,\u201d","author":"Lundberg","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"B32","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1371\/journal.pone.0291928","article-title":"Yield prediction for crops by gradient-based algorithms","volume":"19","author":"Mahesh","year":"2024","journal-title":"PLoS ONE"},{"key":"B33","doi-asserted-by":"publisher","first-page":"456","DOI":"10.2135\/cropsci2005.04-0372","article-title":"A gene-based model to simulate soybean development and yield responses to environment","volume":"46","author":"Messina","year":"2006","journal-title":"Crop Sci"},{"key":"B34","doi-asserted-by":"publisher","first-page":"1819","DOI":"10.1093\/genetics\/157.4.1819","article-title":"Prediction of total genetic value using genome-wide dense marker maps","volume":"157","author":"Meuwissen","year":"2001","journal-title":"Genetics"},{"key":"B35","doi-asserted-by":"publisher","first-page":"691","DOI":"10.1007\/s42106-022-00209-0","article-title":"Potential use of data-driven models to estimate and predict soybean yields at national scale in Brazil","volume":"16","author":"Monteiro","year":"2022","journal-title":"Int. J. Plant Prod"},{"key":"B36","unstructured":"NASA: The power project\n          \n          2025"},{"key":"B37","doi-asserted-by":"publisher","first-page":"4727","DOI":"10.1007\/s10489-021-02662-2","article-title":"Evaluation of interpretability methods for multivariate time series forecasting","volume":"52","author":"Ozyegen","year":"2022","journal-title":"Appl. Intell"},{"key":"B38","doi-asserted-by":"publisher","first-page":"404","DOI":"10.34133\/2019\/5809404","article-title":"Development of optimized phenomic predictors for efficient plant breeding decisions using phenomic-assisted selection in soybean","volume":"58","author":"Parmley","year":"2019","journal-title":"Plant Phen"},{"key":"B39","doi-asserted-by":"publisher","first-page":"110176","DOI":"10.1016\/j.fcr.2025.110176","article-title":"Regional drivers of soybean farmers' yield, seed protein, and oil concentration","volume":"335","author":"Pereyra","year":"2026","journal-title":"Field Crops Res"},{"key":"B40","doi-asserted-by":"publisher","first-page":"1233","DOI":"10.1007\/s11746-999-0099-y","article-title":"Temperature and cultivar effects on soybean seed oil and protein concentrations","volume":"76","author":"Piper","year":"1999","journal-title":"J. Am. Oil Chem. Soc"},{"key":"B41","doi-asserted-by":"publisher","DOI":"10.1145\/2939672.2939778","article-title":"\u201c\u2018Why should i trust you?' Explaining the predictions of any classifier,\u201d","author":"Ribeiro","year":"2016","journal-title":"Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining"},{"key":"B42","doi-asserted-by":"publisher","first-page":"417","DOI":"10.1007\/s10584-016-1601-9","article-title":"Impact of progressive global warming on the global-scale yield of maize and soybean","volume":"134","author":"Rose","year":"2016","journal-title":"Clim. Change"},{"key":"B43","doi-asserted-by":"publisher","first-page":"208","DOI":"10.1016\/j.isprsjprs.2019.12.012","article-title":"Incorporating environmental variables into a MODIS-based crop yield estimation method for United States corn and soybeans through the use of a random forest regression algorithm","volume":"160","author":"Sakamoto","year":"2020","journal-title":"ISPRS J. Photogramm. Remote Sens"},{"key":"B44","doi-asserted-by":"publisher","first-page":"107886","DOI":"10.1016\/j.agrformet.2019.107886","article-title":"Satellite-based soybean yield forecast: integrating machine learning and weather data for improving crop yield prediction in southern Brazil","volume":"284","author":"Schwalbert","year":"2020","journal-title":"Agric. For. Meteorol"},{"key":"B45","doi-asserted-by":"publisher","first-page":"18284","DOI":"10.1038\/s41598-024-69418-z","article-title":"Time series prediction model using LSTM-transformer neural network for mine water inflow","volume":"14","author":"Shi","year":"2024","journal-title":"Sci. Rep"},{"key":"B46","doi-asserted-by":"publisher","first-page":"55","DOI":"10.2135\/cropsci1966.0011183X000600010017x","article-title":"Interception of solar radiation and dry matter production by various soybean planting patterns1","volume":"6","author":"Shibles","year":"1966","journal-title":"Crop Sci"},{"key":"B47","doi-asserted-by":"publisher","first-page":"e0252402","DOI":"10.1371\/journal.pone.0252402","article-title":"Crop yield prediction integrating genotype and weather variables using deep learning","volume":"16","author":"Shook","year":"2021","journal-title":"PLoS ONE"},{"key":"B48","first-page":"3145","article-title":"\u201cLearning important features through propagating activation differences,\u201d","volume-title":"International Conference on Machine Learning","author":"Shrikumar","year":"2017"},{"key":"B49","doi-asserted-by":"publisher","first-page":"199","DOI":"10.1016\/bs.agron.2024.11.003","article-title":"Use of artificial intelligence in soybean breeding and production","volume":"190","author":"Singh","year":"2025","journal-title":"Adv. Agron"},{"key":"B50","doi-asserted-by":"publisher","first-page":"563","DOI":"10.1016\/B978-0-12-817563-7.00011-8","article-title":"Chapter 28-phenomics and machine learning in crop improvement","volume":"15","author":"Singh","year":"2021","journal-title":"Plant Breed. Cult. Dev"},{"key":"B51","first-page":"3319","article-title":"\u201cAxiomatic attribution for deep networks,\u201d","volume-title":"International Conference on Machine Learning","author":"Sundararajan","year":"2017"},{"key":"B52","first-page":"1518","article-title":"Learning global additive explanations for neural nets using model distillation","volume":"1050","author":"Tan","year":"2018","journal-title":"Stat"},{"key":"B53","doi-asserted-by":"publisher","first-page":"e0130855","DOI":"10.1371\/journal.pone.0130855","article-title":"Integrating crop growth models with whole genome prediction through approximate Bayesian computation","volume":"10","author":"Technow","year":"2015","journal-title":"PLoS ONE"},{"key":"B54","doi-asserted-by":"publisher","first-page":"1427086","DOI":"10.3389\/fpls.2024.1427086","article-title":"Temperature and elevated co2 alter soybean seed yield and quality, exhibiting transgenerational effects on seedling emergence and vigor","volume":"15","author":"Thenveettil","year":"2024","journal-title":"Front. Plant Sci"},{"key":"B55","doi-asserted-by":"publisher","first-page":"3393","DOI":"10.1007\/s00122-022-04186-w","article-title":"Genomic selection using random regressions on known and latent environmental covariates","volume":"135","author":"Tolhurst","year":"2022","journal-title":"Theor. Appl.Genet"},{"key":"B56","doi-asserted-by":"publisher","first-page":"1709","DOI":"10.1007\/s00704-022-04341-9","article-title":"Soybean yield prediction by machine learning and climate","volume":"151","author":"Torsoni","year":"2023","journal-title":"Theor. Appl. Climatol"},{"key":"B57","unstructured":"Uniform soybean tests, northern region"},{"key":"B58","unstructured":"Uniform soybean tests, southern region"},{"key":"B59","doi-asserted-by":"publisher","first-page":"e70002","DOI":"10.1002\/tpg2.70002","article-title":"Genomic and phenomic prediction for soybean seed yield, protein, and oil","volume":"18","author":"Van der Laan","year":"2025","journal-title":"Plant Genome"},{"key":"B60","article-title":"\u201cAttention is all you need,\u201d","author":"Vaswani","year":"2017","journal-title":"Advances in Neural Information Processing Systems"},{"key":"B61","doi-asserted-by":"publisher","first-page":"108978","DOI":"10.1016\/j.compag.2024.108978","article-title":"Satellite-based soybean yield prediction in argentina: a comparison between panel regression and deep learning methods","volume":"221","author":"Wang","year":"2024","journal-title":"Comput. Electron. Agric"},{"key":"B62","doi-asserted-by":"publisher","first-page":"e70138","DOI":"10.1002\/csc2.70138","article-title":"Genomic analysis and predictive modeling in the northern uniform soybean tests","volume":"65","author":"Wartha","year":"2025","journal-title":"Crop Sci"},{"key":"B63","doi-asserted-by":"publisher","first-page":"e20084","DOI":"10.1002\/ppj2.20084","article-title":"Phenomic versus genomic prediction\u2013a comparison of prediction accuracies for grain yield in hard winter wheat lines","volume":"6","author":"Winn","year":"2023","journal-title":"Plant Phenome J"},{"key":"B64","unstructured":"\u201cUnified training of universal time series forecasting transformers,\u201d\n          \n          1\n          25\n          \n            \n              Woo\n              G.\n            \n            \n              Liu\n              C.\n            \n            \n              Kumar\n              A.\n            \n            \n              Xiong\n              C.\n            \n            \n              Savarese\n              S.\n            \n            \n              Sahoo\n              D.\n            \n          \n          \n          Proceedings of the 41st International Conference on Machine Learning, ICML'24\n          \n          2024"},{"key":"B65","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/ICOT64290.2024.10936959","article-title":"\u201cCrop yield prediction based on attention-LSTM model,\u201d","volume-title":"2024 International Conference on Orange Technology (ICOT)","author":"Xiong","year":"2024"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1750108\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,26]],"date-time":"2026-02-26T07:37:46Z","timestamp":1772091466000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1750108\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,26]]},"references-count":65,"alternative-id":["10.3389\/frai.2026.1750108"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1750108","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,26]]},"article-number":"1750108"}}