Biological systems collate signals from a large number of molecular binding events, structural or morphological transformations on various scales, and a variety of feedback mechanisms. Uncovering and understanding the physical and chemical driving forces responsible for coordinating such processes remains an important challenge and requires combining insights from various fields, from fundamental equilibrium and non-equilibrium statistical mechanical ideas to ideas from the fields of dimensional reduction and machine learning. In this talk, I will describe two particular examples that highlight some of the progress from the group in developing such new frameworks. First, in the context of cytoskeletal networks, we will show how myosin motor activity combined with actin filament elasticity helps cytoskeletal networks respond adaptively to stresses. A closer look at the non-equilibrium thermodynamics of this system reveals the importance and general role played by energy consumption in controlling this information processing pathway. Next, motivated by advances in the field of active matter where non-equilibrium forcing has been shown to activate new assembly pathways, we will show how non-equilibrium driving in prototypical memory formation models can surprisingly (and in general ways) affect their information processing capabilities. Our results reveal that activity can provide a new and surprisingly general way to dramatically improve the memory and information processing performance without the need for additional interactions or changes in connectivity. Non-equilibrium dynamics can allow these systems to have memory capacity, assembly, or pattern recognition properties, and learning ability, in excess of their corresponding equilibrium counterparts. These results are of significance to a variety of processes involving information storage and retrieval. They are also important for in silico learning and memory forming systems where nonequilibrium dynamics may provide an approach for modulating memory formation.
Self assembly far from equilibrium
There are two related directions of work on non-equilibrium self assembly in our group.
Using the formalism of stochastic thermodynamics, we derive a set of design principles for growing controlled assemblies far from equilibrium. The design principles constrain the set of structures that can be obtained under non-equilibrium conditions and provides intuition for how equilibrium self-assembly landscapes can be modiﬁed under ﬁnite non-equilibrium drive.
M. Nguyen, S. Vaikuntanathan "Dissipation induced transitions in two dimensional elastic membranes", arxiv preprint arXiv:1803.04368, 2019.
Driving forces exerted on a liquid can lead to phase separations not encountered at equilibrium. Understanding the physics leading to such phase separations requires an understanding of non-equilibrium thermodynamics. Using simulations and theory, we study how energy dissipation changes fluctuations in the force the particles in a driven liquid feel. The change in the force fluctuations translate into a change in the diffusion of the particles. A difference in how fast the particles diffuse depending on their local environment can then be used to rationalize the phase separation. We are thus able to make a direct connection between the energy consumed by the system and the phase diagram. The renormalization of force fluctuations by driving forces may be extended to think about how forces in active nematics and other more complex non-equilibrium systems modify their material properties.