Yellow Rust Severity Predictions in Bread and Durum Wheat

This dataset supports research on predicting Yellow Rust severity in wheat using genomic and phenomic data from two biparental populations. It includes data from visual assessments and multispectral UAV imagery for high-throughput phenotyping. The data was used to test parametric, machine learning, and deep learning models.

Use Cases

• Train models to predict Yellow Rust severity from multispectral UAV imagery data.
• Compare the predictability of Yellow Rust using genomic data versus high-throughput phenomic data.
• Benchmark parametric, machine learning, and deep learning model performance on the same wheat population data.
• Analyze Yellow Rust severity patterns across two scoring dates for temporal disease progression.

Strengths

• Data was collected for two distinct wheat species: bread wheat (Triticum aestivum) and durum wheat (Triticum durum).
• Phenotyping was conducted using a multispectral camera on a UAV, enabling high-throughput data collection.
• A diverse set of prediction models, including AI-based frameworks, were evaluated using this data.

Limitations

• The dataset is derived from only two biparental populations, limiting genetic diversity and generalizability.
• Specific sample sizes, feature counts, and data formats are not provided in the input.
• The temporal coverage is limited to two scoring dates, which may not capture the full disease cycle.

Provenance

Source

CIMMYT Harvested Dataverse

Collection Method

Data was gathered from visual Yellow Rust severity scoring and high-throughput phenotyping using a UAV-mounted multispectral camera on two dates; all lines were also genotyped.

Freshness

The dataset was last updated on 2026-02-17.