Genomes to fields 2024 maize genotype by environment prediction competition

TL;DR

This paper introduces a competition to predict maize grain yield using historical and environmental data from 2014 to 2024.

Contribution

The novel contribution is the release of a curated dataset combining genotype, environment, and yield data for maize prediction modeling.

Findings

The dataset includes phenotypic, genotypic, soil, weather, and environmental data from 2014 to 2023.

The dataset was curated with quality control and consistent naming conventions for usability.

The resource supports prediction modeling for maize genotype by environment interactions.

Abstract

The genomes to fields (G2F) 2024 Maize Genotype by Environment (GxE) Prediction Competition challenged participants to develop and submit their best performing models to predict grain yield for the 2024 maize GxE project field trials, using G2F data collected from 2014 to 2023 and other publicly available data.

The G2F Maize GxE Project is a collaborative effort, with all generated data made publicly available. The resource presented here includes the training and test datasets used for the G2F 2024 Maize GxE Prediction Competition. Specifically, data collected from 2014 to 2023 served as the training set to predict grain yield in the 2024 test set. The dataset comprises phenotypic, genotypic, soil, weather, and environmental covariate data, along with metadata describing environments (year-location combinations). It has been curated and lightly filtered for quality control and to ensure consistent naming across years. Competitors also had access to readme files that describe the structure and content of the datasets.