Dataset Information

Yield prediction through integration of genetic, environment, and management data through deep learning.

ABSTRACT: Accurate prediction of the phenotypic outcomes produced by different combinations of genotypes, environments, and management interventions remains a key goal in biology with direct applications to agriculture, research, and conservation. The past decades have seen an expansion of new methods applied toward this goal. Here we predict maize yield using deep neural networks, compare the efficacy of 2 model development methods, and contextualize model performance using conventional linear and machine learning models. We examine the usefulness of incorporating interactions between disparate data types. We find deep learning and best linear unbiased predictor (BLUP) models with interactions had the best overall performance. BLUP models achieved the lowest average error, but deep learning models performed more consistently with similar average error. Optimizing deep neural network submodules for each data type improved model performance relative to optimizing the whole model for all data types at once. Examining the effect of interactions in the best-performing model revealed that including interactions altered the model's sensitivity to weather and management features, including a reduction of the importance scores for timepoints expected to have a limited physiological basis for influencing yield-those at the extreme end of the season, nearly 200 days post planting. Based on these results, deep learning provides a promising avenue for the phenotypic prediction of complex traits in complex environments and a potential mechanism to better understand the influence of environmental and genetic factors.

SUBMITTER: Kick DR

PROVIDER: S-EPMC10085787 | biostudies-literature | 2023 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Yield prediction through integration of genetic, environment, and management data through deep learning.

Kick Daniel R DR Wallace Jason G JG Schnable James C JC Kolkman Judith M JM Alaca Barış B Beissinger Timothy M TM Edwards Jode J Ertl David D Flint-Garcia Sherry S Gage Joseph L JL Hirsch Candice N CN Knoll Joseph E JE de Leon Natalia N Lima Dayane C DC Moreta Danilo E DE Singh Maninder P MP Thompson Addie A Weldekidan Teclemariam T Washburn Jacob D JD

G3 (Bethesda, Md.) 20230401 4

Accurate prediction of the phenotypic outcomes produced by different combinations of genotypes, environments, and management interventions remains a key goal in biology with direct applications to agriculture, research, and conservation. The past decades have seen an expansion of new methods applied toward this goal. Here we predict maize yield using deep neural networks, compare the efficacy of 2 model development methods, and contextualize model performance using conventional linear and machin ...[more]

PMID: 36625555

Similar Datasets

Project description:Specifically, Iraq is threatened in its second-largest northern city, Mosul, by the collapse of the Mosul Dam due to problems at the root of the dam, causing a wave of floods that will cause massive loss of life, resources, and public property. The objective of this study is to effectively monitor the level of dam water by predicting the level of water held by the dam In anticipation of achieving flood stage and breaking the dam, and supporting its behavior through formation 14-day time series data to predict seven days later. Used six deep learning models (deep neural network (DNN), convolutional neural network (CNN), convolutional neural network long short-term memory (CNN-LSTM), CNN-LSTM-Skip and CNN-LSTM Skip Attention) that models were trained to predict the water level in the dam; these levels of being under surveillance and prepared In case of danger, alert people to potential flood threats depending on the dam's water level. These time series were created from the actual data sets of the dam; it's a fundamental historical reading for 13 years (1993-2006) of the water level stored in the Mosul dam and was adopted in coordination with the Iraqi Ministry of Water Resources and the Centre for Research on Dams and Water Resources at Mosul University. The methodology applied in this study shows the model's performance efficiency and the prediction results' low error rate. It also compares those practical results to determine and adopt the performance-efficient model and a lower error rate. The comparison of these results proved the accuracy of its results, and superior to the CNN-LSTM model, it has the highest ability to perform through high accuracy with MAE = 0.087153 and time steps = 0 s 196 ms/step and loss = 0.00067. The current study demonstrated the ability to predict the water level in Mosul Dam, which suffers from foundation problems and may collapse in the future. Therefore, the water level in the dam must be monitored accurately. It also aims to test the effectiveness of the six models proposed in this study after evaluating their performance and applying the prediction process within a scenario to obtain predictive values after 14 days. The results showed the practical effectiveness of the hybrid CNN-LSTM model in correctly and accurately obtaining predictive values within the integrated framework of the required scenario. The study concluded that it is possible to enhance the ability to monitor and identify the potential risk of Mosul Dam at an early stage, and it also allows for proactive crisis management and sound decision-making, thus mitigating the adverse effects of crises on public safety and infrastructure and reducing human losses in areas along the Tigris River.

Dataset Information

Yield prediction through integration of genetic, environment, and management data through deep learning.

Publications

Yield prediction through integration of genetic, environment, and management data through deep learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets