Phenotyping grapevine resistance to downy mildew: deep learning as a promising tool to assess sporulation and necrosis.

Summary

The detailed study, assisted by imagery, of plant leaf symptoms during compatible or incompatible interaction with pathogenic microorganisms allows the description and measurement of the dynamics of these interactions in space and time. This step is necessary for the identification and understanding of resistance or tolerance mechanisms in plants and is therefore a major challenge in plant improvement.
This thesis project aims to develop software tools to exploit the high-throughput phenotyping platforms TPMP in Toulouse and VEGOIA in Colmar to better understand the dynamics of the appearance and to quantify foliar symptoms caused by three pathogens commonly found on Arabidopsis thaliana (Pseudomonas, Xanthomonas, Sclerotinia) and three pathogens encountered on grapevine (downy mildew, powdery mildew and black rot). The objective is to develop and deploy ad-hoc deep-learning tools that can address in the most generic way possible the analysis challenges posed by the different pathosystems and the different scales of observation. The study material will be parts of organs (grapevine leaf discs, VEGOIA), isolated whole plants (Arabidopsis, TPMP) and, later, larger isolated plants (Grapevine, TPMP) with the prospect of adapting the analytical tools to plants in natural conditions (Grapevine, Colmar).
The research is intended to develop deep-learning algorithms that make the best use of the images generated by the dedicated platforms. The supervision of this thesis will be twofold with a biologist thesis director (Nemo Peeters) and a thesis director specialized in deep-learning (David Rousseau).

Contact

• Felicia Maviane-Macia
• Linkedin

Publications

Felicià Maviane Macia, Tyrone Possamai, Marie-Annick Dorne, Marie-Céline Lacombe, Eric Duchêne, et al.. Phenotyping grapevine resistance to downy mildew: deep learning as a promising tool to assess sporulation and necrosis. Plant Methods, 2024, 20 (1), pp.90. ⟨10.1186/s13007-024-01220-4⟩. ⟨hal-04630893⟩