Return to Colloquia & Seminar listing
Regularization of collapse in cellular dynamics: connecting microscopic and macroscopic scales
Mathematical Biology| Speaker: | Pavel Lushnikov, U New Mexico |
| Location: | 2112 MSB |
| Start time: | Mon, Feb 14 2011, 3:10PM |
Description
Dynamics of biological cells and bacteria at the level of individual
cell (microscopic scale) is usually complicated and involves
self-propelled random walk,
motion in chemotactic gradient or directed motion with quasi-periodic
reversals.
It makes the task of the modeling of the dynamics of large ensembles of
cells quite challenging. We used the microscopic motion of cells to
derive the macroscopically
averaged equations for the dynamics of the cellular density (coupled
with the
dynamics of chemoattractant if relevant) in two particular cases. First
case is the dynamics of the cells with strongly fluctuating shape with
their centers of mass experiencing the random walk relevant e.g. for
the the dynamics of
Dicty amoeba or mesenchymal cells. Second case is the dynamics of
Myxobacteria which move with near constant speed and experience periodic
reversals of the direction of their motion.
In all cases cells interacts through the excluded volume constraint
which does not allow cells to penetrate into each other. The resulting
partial differential equations (PDEs) for the cellular density have the
form of the
nonlinear diffusion equation (although the nonlinear diffusion
coefficients are different in these two cases).
The derivation of these macroscopic equations is the loose analog of
the derivation of the Navier-Stokes equations from the dynamics of
individual molecules (except that the individual cell dynamics is highly
overdamped
and includes different types of self-propelled motion while the
molecular dynamics is Newtonian). We show that perturbation theory
breaks for quite small cellular densities and instead non-perturbative
approaches were
developed. The resulting PDEs have no fitting parameters and depend
only on the parameters of the individual cellular motion which can be
easily measured in experiments. We demonstrated very good quantitative
agreement
between large scale Monte Carlo simulations of the microscopic
stochastic dynamics of the ensembles of cells and the solutions of the
derived PDEs.
