At the heart of DREAM is a series of experiments designed to acquire data on the dynamics of photosynthesis regulation to be modelled and interpreted with theoretical techniques.
The theoretical models should be able to link the recorded responses to specific physiological states and thus provide protocols tailored to the organism’s needs: clear instructions to treat or support the organism with the right resources.
First, we will implement a black-box model based only on collected data interpreted through AI. This technique has no physical or biological content, as we do not need to know the molecular mechanisms behind the phenomena. However, since photosynthesis has 100 years of research behind it, we can exploit all available knowledge about its processes and molecules.
The challenge for us is refining the black-box model by introducing some of this information, considering the different experimental conditions. Indeed, unlike DREAM, photosynthesis research often requires time-consuming normalizing protocols (i.e., putting the organism in the dark for a long time) and does not focus on real-time measurements in natural light.
The resulting model should guide the design of plant-specific monitoring protocols to reduce the demand for water, nutrients, and pesticides. In DREAM, we will validate the protocols in controlled environments on three organisms: the green alga C. reinhardtii, the model plant A. thaliana, and the tomato.