AI to understand plant resilience to extreme environments | Engineering Techniques

From plants collected in the Atacama Desert, scientists have used a new technique to highlight their mechanisms for adapting to extreme conditions. Called predictive metabolomics, it consists of studying the chemical composition of plants and using artificial intelligence.

The Atacama Desert, located in South America, is considered one of the driest regions on earth. Several varieties of endemic plants are still present on site. After collecting several species growing between 2,400 and 4,500 meters above sea level, scientists from INRAE, the University of Bordeaux and the Pontifical Catholic University of Santiago in Chile have managed to identify common molecular markers allowing to understand the mechanisms of resilience of these plants in the face of an extreme environment. The researchers used an innovative approach using artificial intelligence. The results of their work are detailed in the review New Phytologist.

Thomas Dussarrat, PhD student on this research project, harvesting plants in the Atacama Desert. Credit: Isabel Mujica

“We used a technique called predictive metabolomics, explains Pierre Pétriacq, teacher-researcher at the University of Bordeaux and at INRAE. It consists of the study of the metabolome, that is to say all the small molecules present in a living organism, such as sugars, amino acids, antioxidants. The originality of our approach then consisted in coupling this technique with artificial intelligence. »

In concrete terms, 24 species were analyzed and made it possible to study approximately 5,000 different metabolic signals. Thanks to the use of machine leaning, a statistical approach based on the training of a machine learning algorithm, the scientists succeeded in highlighting 39 molecules common to all these species, making it possible to predict the environment in which grows the plant with 79% accuracy. Among these 39 molecules, starch was identified as the best predictive chemical compound.

“The prediction rate is very high, we did not expect it, analyzes the researcher. With phenotypic and physiological data such as plant size, yield or photosynthesis, it is not possible to achieve such high precision. In addition, our study was carried out on multi-species plants, whereas often researchers are interested in only one species in particular. » From this result, the researchers managed to make the link between the expression of these 39 molecules with different environmental stress factors such as water stress, very high light intensity, high salinity in the soil, etc.

Reduce development times for new plant species

Since these 39 chemical compounds are also present in plant species grown in other climates, such as corn, peas, tomatoes or sunflowers, this new approach could also help to better target plants whose metabolism is the more likely to withstand environmental stresses. “This is a research hypothesis that will have to be tested, says Pierre Pétriacq. The idea would be, for example, to find varieties more quickly whose metabolism is close to that observed in the plants of the Atacama desert. This could reduce the time needed to adapt cultivated plants to the constraints of climate change. Today, it takes ten years for scientists and producers to develop a new plant species that meets specific environmental criteria such as frost resistance. »

Another avenue of application, this time on the agronomic level. Through the use of predictive metabolomics which combines the study of chemical composition and the use of AI, the researchers plan to try to predict the behavior of plants according to different cultivation practices. “We could imagine scenarios where we grow plants with more or less virtuous practices, such as reducing phytosanitary products, then try, for example, to identify associations of species that are part of an agroecological approach”adds the researcher.

Finally, a last line of work could apply not to our planet, but to that of Mars. The soil of the Atacama is indeed the one that most closely resembles Martian soil. The researchers of this project are in contact with scientists from ESA and NASA and have initiated a reflection which could lead them to use this new approach to identify the plants potentially likely to grow on Mars. Of course, the idea is not to try to cultivate it on this planet, but to carry out basic research work.

We would like to thank the author of this write-up for this remarkable web content

AI to understand plant resilience to extreme environments | Engineering Techniques

Visit our social media profiles as well as other related pages