Harini Rangarajan

My doctoral dissertation titled “Exploring the Root Phenome: Simulation Modeling with a Functional Structural Plant Model “with Prof.  Jonathan P Lynch at the Pennsylvania State University demonstrates the use of in-silico methods to understand the root phenome. In my research, I used root modelling tools to simulate root systems in environments varying in nutrient availability to evaluate their functional performance in terms of plant productivity. My research demonstrated the application of multi-objective optimization sub-field of Artificial Intelligence to search through the large decision space presented by the varying root phenotypic traits (phenes) and the environment to identify optimal root phenotypes. I used unsupervised machine learning techniques such as self-organizing maps to cluster and analyze the resulting high -dimensional data obtained from the optimization studies, as well as supervised machine learning techniques such as random forests to analyze data in a different study to evaluate and determine most informative measurements of root architecture

Studies of the various processes that occur in the soils, in the roots at the various scales, soil/ plant interactions and their effect on nutrient uptake, plant growth and productivity interest me.  The weaving of all the mechanistic processes involved at desirable resolution is computationally challenging but holds promising solutions towards food security and is an effort I aim to be a part of.