Abstract
As sessile organisms, plants are continuously exposed to several abiotic stress factors such as extreme temperatures, drought, salinity and metal toxicity, which limit growth and crop productivity worldwide. For this reason, plant-environment interactions have become one of the most active research fields in plant studies to counteract the effects of these stress factors on agricultural yields. The emerging use of biostimulants to amicably improve plant growth and performance has been at the heart of modern agriculture research to improve crop quality while maintaining organic crop production, cost- and time-efficient everlasting systems. However, the mechanisms by which biostimulants enhance growth, improve nutrient uptake, and increased resistance and tolerance to abiotic stresses are still poorly understood. In this context, the current work is an untargeted metabolomics study aimed to unravel the mechanisms underlying the activity of biostimulants in abiotic stress amelioration. In addition, the current study goes beyond the standard machine learning (ML) linear projection methods for data analysis and explores the more advanced ML nonlinear regression algorithms to capture the complex and highly multicollinear data generated in untargeted metabolomics studies
M.Sc. (Biochemistry)