dr. Avishek Das
CV / Biography
– What makes a collection of chemical reactions into a living organism?
– How do biological cells perceive and manipulate their noisy world?
– How does chemistry and diffusion come together to create _meaningful_ spatiotemporal patterns?
These are the core questions that fuel my research in the information processing of soft and biological matter. I strive to answer these through fundamental theory in statistical physics and information theory, through the development of novel numerical algorithms, and by collaboration with experimentalists who characterize and design intelligent matter. We only fully understand how something works if we can make a working prototype from scratch in the lab!
Currently in my postdoctoral research in the group of Prof. Pieter Rein ten Wolde at AMOLF, I am exploring the similarity between biologically evolved motile bacteria and a hypothetical microrobot that optimally climbs up the gradient of a chemical signal. Using the framework of information theory, we are developing novel theoretical, numerical and experimental techniques to quantify the directional information processing in general biochemical networks, and specifically for stochastic navigators.
I did my PhD in Physical (Theoretical) Chemistry with Prof. David Limmer at the University of California Berkeley. There I designed a novel numerical paradigm, called Variational Path Sampling, for simulating and designing nonequilibrium fluctuations of soft matter. My research synthesized tools from molecular dynamics simulations, large deviation theory in probability, stochastic thermodynamics, and reinforcement learning. We produced several efficient numerical algorithms to sample rare fluctuations, infer reaction rate constants, design self-assembly and create life-life properties in colloids.
In the future, I would like to address how the two fields come together to make soft matter intelligent. Specifically, I would like to study information processing in systems where both reactive and diffusive components dynamically carry information. Such systems are ubiquitous in biology and are only starting to be designed _in vitro_. In such systems, the quantification and optimization of the amount of processed information require new theoretical and numerical innovations, due to the wide range of timescales involved. I want to work towards developing those tools, apply them to the experimental characterization of biological systems, and discover the key components to create _Droplet Computers_: which is a synthetic matter that can compute using both reactions and diffusion.
A favorite activity of mine, besides pondering about the origin of life, is teaching. I have experience as a teaching assistant for several graduate-level physical chemistry courses. I like teaching so much that, after my teaching assistantship was over, I co-founded, designed and taught a yearly mathematics bootcamp for physical chemistry PhD students at Berkeley.
Research Interests
Information processing in soft matter, Biophysics of cellular motility, Nonequilibrium self-organization in colloids, Rare event sampling, Numerical algorithms for statistical physics
