| |
Noise propagation in biochemical reaction network
Biochemical reactions occurring inside a tiny living cell, say a protein phosphorylation by a kinase, are fundamentally different than reactions carried out in a test tube. The differences are twofold: a) the size of reaction chamber (the cell volume) in case of a living cell is way too small compared to a test tube, typically in the femto-litter (fL) range, b) the concentrations of various reactants inside the cell vary in the micro to nano molar range. Because of the tiny volume of a living cell, the numbers of molecules of various chemical species inside the cell are quite small (in the range of 1 to 100000 molecules per cell on average). Owing to the low numbers of species, the effect of molecular fluctuations (chemical noise) of reactants is significant on product for a chemical reaction occurring inside a living cell.
Because of the chemical noise, a population of genetically identical living cells show significant amount of cell-to-cell variability in characteristics/cellular content even under in same environmental conditions. For unicellular organism the main source of variability is the gene expression noise: noise associated with the production of a protein. Owing to the bigger cell volume of a multicellular organism, it was thought that the cell-to-cell variability in multicellular organism would be less compared to unicellular organism, but recent single cell data contradicts the presumption. In fact single cell experiments on mammalian cells indicate that the non-genetic sources of noise play a major role in creating cellular diversity: particularly how noise is propagated along a chemical reaction network.
Our goal in the lab is to understand the dynamics of chemical noise propagation through a biochemical reaction network consisting of various kinds of feedback and feed-forward regulatory motifs. Using small regulatory motifs we build small regulatory modules and applying statistical mechanics, stochastic calculus, non-liner dynamics and stochastic simulations, we try to understand how the regulatory motifs modulate the outcome of chemical noise.
|
Finding motif/module from noisy data
Based on the knowledge we gain, described in the earlier section, our plan is to go backwards. Our goal is to formulate statistical mechanical methods to predict the regulatory connections among the various species, given the distribution of various species (say, proteins, mRNAs etc.). We plan to compare results from our theoretical models with single cell quantitative data from literature. |
Stochastic simulation method for coupled chemical reaction network
Numerical simulation of biochemical reaction network is one of the most important tools that researchers use enormously to explore the dynamics of noise in a network. The stochastic simulation algorithm (SSA) proposed by D. T. Gillespie, commonly known as Gillespie’s algorithm, is the most widely used method to simulate coupled chemical reaction systems. Although it usefulness is widely accepted, but the computational cost is heavy for a large network with disparate timescales and species abundances. This can be the case in enzyme-catalyzed reactions where the association and dissociation of the enzyme and substrate occur much more frequently than the product producing reaction and abundance of enzyme is quite less than that of the substrate. We plan to use stochastic quasi steady state approximation of chemical master equation to improve upon SSA such that the computational cost of simulation reduces without the accuracy of simulation. |
Heat conduction in low dimensional lattice
Establishing macroscopic laws for various properties of a system from microscopic principles is one of the most important problems in science. Verification of Fourier’s law of heat conduction in finite dimensional systems has been major research interest recently. We work on thermal conduction in 2-dimensional lattices with different kinds of interacting potentials to find out macroscopic description of heat transport in finite dimensional systems.
Final Project Report: UGC MRP (2015)
|
|
