Soybean Root Nodulation Traits Quantified by YOLO Pipeline for GWAS

A YOLO-based deep learning pipeline quantified 24 nodulation traits from root images of 21-day-old hydroponically grown soybean plants. The model achieved a precision of 0.94 and processed 50 images in 37 seconds, representing a ~227-fold efficiency gain over manual scoring. The dataset, published by Kuber Shivashakarappa in April 2026, was used in a genome-wide association study that identified 50 significant SNPs.

Use Cases

• Conducting genome-wide association studies (GWAS) based on automated nodulation trait measurements.
• Training or benchmarking object detection models for plant root phenotyping based on the described YOLO pipeline.
• Analyzing genetic determinants of symbiotic nitrogen fixation based on quantified nodule count, dimensions, and spatial distribution.
• Developing scalable phenotyping workflows for large-scale genetic studies based on the reported efficiency gains.

Strengths

• Model performance is quantified with a precision of 0.94, recall of 0.95, and F1 score of 0.94 for nodule detection.
• The pipeline measures 24 distinct nodulation-related traits, enabling detailed phenotypic analysis.
• Processing efficiency is documented as a ~227-fold improvement, taking 37 seconds for 50 images versus ~2 hours 20 minutes manually.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale statistical analysis.
• The dataset is very small (18.5 KB), suggesting it may contain summary statistics or a limited sample rather than raw image data.

Provenance

Source

figshare

Collection Method

Traits were extracted from root images using a YOLO-based deep learning phenotyping pipeline.

Freshness

Last updated 2026-04-21 05:25:35; freshness should be verified.