Accelerated Haustoria Segmentation Enables Rapid Gene Function Analysis in Cereal-Powdery Mildew Pathosystems
Stefanie Lück1, Deniz Demirhan1, Laura Agsten1
1Leibniz Institute of Plant Genetics and Crop Plant Research (IPK), 06466 Seeland OT Gatersleben, Germany.
Abstract:
Reliable, high-throughput quantification of early fungal infection events is crucial for gene function studies, but it remains labor-intensive. We report an openly available pipeline that automates the detection of β-glucuronidase (GUS)-stained epidermal cells and the intracellular haustoria formed by powdery mildew on barley and wheat leaves. Whole-slide images are captured with a commercial scanner, focus-projected, tiled, and analyzed by deep-learning models trained on expertly annotated datasets. A You Only Look Once (YOLO) network identifies GUS-positive cells, and a companion segmentation model pinpoints haustoria within each cell; automatic focus-layer selection preserves fine structural detail. The workflow runs in minutes per slide on a single workstation and maintains near-perfect agreement with manual counts in both barley and wheat, demonstrating robust cross-species transferability. By delivering single-cell readouts with minimal user input, the pipeline enables rapid functional validation screens and supports large-scale phenotyping of cereal-powdery mildew interactions. [Formula: see text] Copyright © 2025 The Author(s). This is an open access article distributed under the CC BY-NC-ND 4.0 International license.
Insights
A new pipeline automates the detection of fungal infections in plants using deep learning. This tool accurately quantifies powdery mildew in barley and wheat, speeding up gene function studies.
Area of Science:
- Plant Pathology
- Computational Biology
- Genetics
Background:
- Quantifying early fungal infections in plants is essential for genetic studies but is currently labor-intensive.
- Powdery mildew infection in cereals like barley and wheat requires efficient analysis methods.
Purpose of the Study:
- To develop and validate an automated pipeline for high-throughput quantification of fungal infection events.
- To enable rapid functional validation screens and large-scale phenotyping of cereal-powdery mildew interactions.
Main Methods:
- An openly available pipeline was developed using deep learning models (You Only Look Once network and segmentation model).
- The pipeline analyzes whole-slide images of GUS-stained epidermal cells and intracellular haustoria in barley and wheat leaves.
- Automatic focus-layer selection was implemented to preserve fine structural details.
Main Results:
- The pipeline accurately detects β-glucuronidase (GUS)-stained cells and haustoria in both barley and wheat.
- Automated counts showed near-perfect agreement with manual counts, demonstrating robust cross-species transferability.
- The workflow processes slides in minutes per workstation with minimal user input.
Conclusions:
- The developed pipeline significantly reduces labor intensity in quantifying fungal infections.
- This automated system enables rapid functional validation and large-scale phenotyping of plant-fungal interactions.
- The pipeline supports efficient research in cereal-powdery mildew interactions.
