Deep learning and Deep characterization of Peptides for the control of plant nutrition – DeepPep

Nitrogen (N) and Phosphorus (P) are two major nutrients sustaining food production and consequently human population development. However, N and P provision in agricultural systems have detrimental effects on ecosystems mainly because these nutrients tend to leak in the biosphere due to a deficient match in between plant capacity to use them, and fertilizer provision. These leaks trigger major water pollution events as well as are important atmospheric modifications responsible for climate change.
This is why understanding and controlling plant N and P nutrition may thus have crucial consequences for sustainable human development.We recently obtained molecular insights showing that a way to control and improve NUE and PUE may lie in the intimate control of N signaling over P signaling.The objective of the NutriStim project is thus to acquire fundamental knowledge to control these molecular pathways by using signaling peptides.
For this, we will engage in an unprecedented “pan-peptidomic” approach to identify natural peptides being controlled by N and P signals in 3 plant species.
In parallel, we will develop a deep learning (Artificial Intelligence [AI]) model having the ambition to predict synthetic peptides able to interact with a given target protein. These target proteins will be selected in the NxP signaling pathway.
These natural and synthetic peptides will then be applied to plants to record their effect on developmental and transcriptomic responses. For the most promising ones, in depth functional genomics will be engaged. Ultimately, the DeepPep team made of 5 highly complementary labs (at CNRS, Univ Montpellier, INRIA, INRAE) will provide in depth characterization and understanding of the regulatory peptides at play in plants, opening direct perspective in biotechnological applications in the field of biostimulants and AI drug design.