This project is a collaboration between the Endres and Pawar labs.
There has been a lot of interest in understanding the diversity and stability of microbial communities. Despite significant progress in specific areas of microbial research, fundamental thermodynamic principles determining the evolution and adaptation of complex microbial communities are still very little understood. In this project, we will describe microbial communities as networks of interacting species, and explore different naturally occurring network topologies, including scale-free, small-world, clustered, hierarchically-modular and examples taken from databases . Next, we will study their entropy production as a measure of energy turnover and vitality. In particular, a key novelty will be studying the evolution toward minimal and maximal entropy production networks using genetic algorithms and recurrent neural networks to capture mature near and young far-from-equilibrium microbial communities. Resulting networks will be characterised in terms of metrics including connectivity, mean shortest path length, and degree distribution.
The work is a natural extension of previous and ongoing work in Prof Endres' biological physics group, some done in collaboration with co-supervisor Dr. Pawar with expertise in individual and community-level metabolism. These works include the emergence of cooperativity in biofilms , the assembly and diversity of microbial ecosystems , and spatial pattern formation in bacterial colonies. Often we make a link between mechanistic modelling and machine learning in my group. We anticipate that identified thermodynamic principles will be widely applicable to systems ranging from chemical reaction networks and microbial communities to economics.