Soybean Phenotypic and Genotypic Data for Genomic Selection Model Testing

Phenotypic and genotypic data from a soybean variety development program, collected across eight locations during the 2023 and 2024 growing seasons. The dataset includes results from testing six genomic selection models and four training population optimization methods. It was created by Leonardo De Azevedo Peixoto and is licensed under CC-BY-4.0.

Use Cases

• Compare performance of genomic prediction models (rrBLUP, Bayes A, Bayes B, RKHS, random forest, SVM) based on seed yield, protein, oil, and maturity traits.
• Optimize training population structure and size for genomic selection based on the four methods (RS, MGRS, ERS, GA) and ten size percentages evaluated.
• Implement multitrait selection strategies (Rank-index, Smith-Hazel index) for soybean breeding programs based on the described selection index strategies.

Strengths

• Data includes experiments planted in a randomized block design with two replications, suggesting controlled experimental conditions.
• Six distinct genomic selection models and four training population optimization methods were tested, providing a comparative framework.
• The dataset includes R scripts, which likely facilitates reproducibility and analysis.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset is 1.2 MB, which is a small scale and may limit the complexity of analyses.

Provenance

Source

Soybean variety development program.

Collection Method

Experiments planted in a randomized block design with two replications at eight locations.

Time Range

2023 and 2024 growing seasons.

Freshness

Last updated 2026-05-05 20:30:20; freshness should be verified.

Geography

Eight locations (specific locations unknown).