Open digital phenotyping to accelerate the creation of aphid-resistant plant varieties

Context and key challenges

Although the post-genomic revolution has produced a suite of exceptional tools for the molecular description of plant-insect interactions, the phenotyping of plant genetic resources has lagged behind due to the sheer diversity of traits of interest and insect-related specificities in existence. Meanwhile, aphids currently head the list of major insect pests arousing concern because of the ease with which they act as vectors for numerous viral diseases. Chief among the major insect pests that are cause for concern are aphids, due to their ability to transmit numerous viral diseases.

The task of phenotyping plant resistance to aphids is both slow and labor-intensive, calling for a vast number of measurements to ensure that calculations are accurate.

The GRATITUDE consortium has been formed to resolve this issue through the development of innovative image-capture tools that will be operate in conjunction with an analysis pipeline that employs computer vision. This system will make it possible to:

1.  quantify aphids according to their developmental stage
2.  characterize aphid behaviors (dispersion, preference, feeding, etc.) on the plant.

The ‘Open by Design’ system that the project plans to build will establish a standard that is both open and scalable. The accessibility of the system’s hardware, data and analysis pipeline set it apart from the majority of commercial digital phenotyping tools, where closed systems are the norm and vendor lock-in is often imposed.

Goals and methodology

The GRATITUDE pathway establishes a sequence of key actions:

• The first action will be to hold a scientific residential workshop. This will establish the state of the art for the application of deep learning and neural networks to plant-insect interactions and will identify the most promising models and approaches;
• Second, two Masters 2 internships co-supervised by the project partners will allow the selected methods and biological materials to be tested. During these internships a series of training sessions will be held on the use of the chosen methods, coordination of data collection, and analysis of the data generated by the project;
• Last, a further residential workshop will be held to write up the results in the form of a joint scientific publication. Results will be shared with the INRAE community, notably within the BAPOA and ModStatSAP networks.

By mapping the genetic determinants of plant resistance to aphids (and other pests), this project will form the first step in developing accurate and predictive high-throughput digital phenotyping tools and, ultimately, will improve host-plant resistance by optimizing genetic selection.

Contact - Coordination:

Participating INRAE units and external partners

INRAE units

External partners