Developing Deep Learning Models And Tools To Score Plant Cell Death And Disease Lesion Severity


Plant pathogens are a major threat to global crop production, incidence of crop disease is increasing, and global climate change and agriculture expands the geographical range of pathogens threatening food supply further. Monitoring effects of plant pathogens involves assessment of immune responses or disease by scoring severity of cell death or measuring size of disease lesions on the plant. Current scoring methods are based on photos of samples (with visible and/or UV light), image processing (e.g with Photoshop), layout and manual scoring by comparison with an external reference image, a severe bottleneck to many important experiments. To deal with the threat of plant pathogens we need tools that can perform fast, high-throughput cell death/lesion assessment in the lab and field.

The overall project aim is to develop a novel machine vision tool using deep learning models like Convolutional Neural Networks that can score images automatically. You will be able to develop these in Python and R frameworks like PyTorch and Tensorflow. Once trained and optimised the models can be embedded in low-power easy to use hardware and software for deployment in lab and field.

You will enjoy this project if you have a background in computer science or informatics and an interest in software and hardware aspects of technology. By the end of the project you will have developed deep expertise in software engineering, data science, statistics, and plant pathology, a combination of skills that is of high value in the academic and industrial job market. You will be supported by extensive interactions between biological and computationally focussed research groups.


Kristianingsih, R & MacLean D (2021) Accurate plant pathogen effector protein classification ab initio with deepredeff: an ensemble of convolutional neural networks. BMC Bioinformatics 22: 372

De la Concepcion JC, Franceschetti M, MacLean D, Terauchi R, Kamoun S and Banfield MJ (2019) Protein engineering expands the effector recognition profile of a rice NLR immune receptor. eLife 2019;8:e47713

MacLean (2020) redpatch – a package to find disease lesions in plant leaf images doi:10.5281/zenodo.3965768