pyRootHair: Machine learning accelerated software for high-throughput phenotyping of plant root hair traits

Area of Science:

• Plant Biology
• Genetics
• Computational Biology

Background:

• Root hairs are crucial for plant water and nutrient uptake.
• Manual quantification of root hair traits is time-consuming and low-throughput.
• Automated phenotyping tools are needed for rapid genetic advancements.

Purpose of the Study:

• To develop and validate pyRootHair, an AI-powered software for high-throughput root hair trait extraction.
• To enable rapid screening of plant germplasm for root hair morphology variation.

Main Methods:

• Development of pyRootHair, an AI application for automated root hair trait extraction from microscope images.
• Batch processing of over 600 images per hour without manual intervention.
• Deployment on diverse wheat cultivars (Triticum aestivum, Triticum turgidum ssp. durum) and other plant species.

Main Results:

• pyRootHair successfully automates root hair trait extraction with high throughput.
• Uncovered significant, previously unresolved variation in root hair traits across wheat cultivars.
• Identified two distinct root hair shape categories and correlations between traits.
• Demonstrated applicability across multiple plant species (oat, rice, teff, tomato).

Conclusions:

• pyRootHair facilitates rapid, high-resolution phenotyping of root hair morphology.
• Enables large-scale screening of plant germplasm for genetic studies.
• Supports investigation into the genetic control of root hair traits and their effect on plant performance.