Soybean Yield Genomic Prediction Data for 1,789 Genotypes

Arkansas and Missouri soybean genotypes from maturity groups III to V were evaluated across 10 environments during the 2023 and 2024 growing seasons. Rafael Goncalves Marmo created this dataset to support a classification-based genomic prediction framework for identifying high-yielding genotypes. The dataset likely contains genomic predictors and yield performance classes derived from SoySNP3K BeadChip markers.

Use Cases

• Training classification models for yield prediction based on genomic marker data.
• Comparing the performance of GLMNet and Random Forest models for genomic prediction.
• Identifying high-yielding and stable soybean genotypes based on a selection index (MSI).
• Analyzing the relationship between genomic data and yield consistency across multiple environments.

Strengths

• Includes 1,789 soybean genotypes, providing a substantial sample size.
• Data spans 10 environments (year-location combinations) across two states.
• The classification models achieved macro-averaged balanced accuracies of 0.84 and 0.83.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is 10.4 KB, indicating a very limited scope.

Provenance

Source

figshare

Collection Method

Genomic Best Linear Unbiased Predictors (GBLUPs) were calculated from SoySNP3K BeadChip markers and yield data from field trials.

Time Range

2023-2024 growing seasons

Freshness

Last updated 2026-04-29 05:41:17; freshness should be verified.

Geography

Arkansas and Missouri, USA