Thema der Dissertation:
Non-Markovian dynamics of cells and proteins
Non-Markovian dynamics of cells and proteins
Abstract: The generalized Langevin equation (GLE) provides a unifying framework to describe the stochastic dynamics of coarse-grained observables in complex systems. We employ the GLE to investigate how memory effects and energy barriers shape diffusive behavior across physical, biological, and active systems. The interplay of these effects is not well understood. Thus, we disentangle the influences of energy barriers and memory effects on subdiffusive dynamics. We find that memory effects dominate the subdiffusive dynamics for overdamped systems up to a potential-dependent relaxation time. We give expressions for all relaxation time scales and derive the mean squared displacement (MSD) for general multi-exponential memory in the GLE. Further, we find that the friction kernels extracted from published fast-folding protein molecular dynamics (MD) simulations exhibit a hierarchical structure that resembles a truncated power law. By comparing GLE simulations and theory, we find that the characteristic protein-folding subdiffusion originates mostly from memory and not from the shape of the free energy landscape. Markovian models fail to capture the subdiffusive dynamics. Extracting the GLE parameters of moving Chlamydomonas reinhardtii (CR) algae cells reveals individual differences between single cells in their motility patterns, which allows for single-cell classification. Due to the generality of the GLE framework, it is readily applicable to any kind of cell motion and to arbitrary observables. Applying the GLE framework to human breast-cancer cells, we find that the motion parameters of cells with identical DNA intrinsically vary over two orders of magnitude. This highlights how large phenotypic differences can arise even among cells with identical genetic information.
Time & Location
Apr 08, 2026 | 04:00 PM
Hörsaal A (1.3.14)
(Fachbereich Physik, Arnimallee 14, 14195 Berlin)