Modeling and simulations of biopolymers and synthetic polymers
31st January 2018, Wednesday 12:00AM Seminar Hall, Dept of Chemical Engineering.
Abstract
Processes that span across multiple length scales and timescales are common in nature, and the multiscale nature often limits our knowledge of biological systems. Systematic models based on a strong fundamental basis serve as powerful tools to probe such systems in areas that lie outside current experimental limits, to complement our experimental knowledge about these systems. Such models can also serve as guides to design synthetic mimics with useful industrial applications such as DNA sequencing, polymer characterization, etc. One can typicallly make progress at constructing these models by making approximations such as separation of timescales and/or length scales. On the other hand, certain processes, especially in biology, are inherently multiscale in nature and hence demand alternative treatments that do not invoke these approximations. In this talk, I will discuss a computational model of actin filaments as a demonstration of modeling a process of the latter kind and a model of polymer translocation as an example of the former.
Actin filaments continually assemble and disassemble in the cell cytoplasm, during which they age as the nucleotides bound to newly added actin sub-units hydrolyze from ATP to an intermediate ADP-Pi, followed by a slower release of Pi to form ADP. This results into a spatial distribution of all three states of the bound nucleotide along the growing filament. It is not fully understood whether ATP hydrolysis and Pi release in a growing actin filament are spatially cooperative or purely random. I will discuss the multiscale nature of this problem, and describe a bottom-up coarse-grained modeling strategy that helps us probe in the mechanism of ATP hydrolysis in these filaments. The modeling approach can be extended to other cytoskeletal problems of multiscale nature such as dynamic instability in microtubules and dynamics of filopodia formation.
Sequencing DNA by electrophoretically translocating it single-file across a nanopore can potentially serve as a low cost high throughput alternative to existing techniques. Although significant progress has been made in understanding the rich phenomenology in volved in the process since it was proposed over two decades ago, achieving a desirable level of control over the process is a fundamental challenge. I will discuss the relevant phenomenology along with the multiple timescales involved, and describe a model based on the Fokker-Planck formalism that invokes the approximation of separation of timescales. In the end, I will discuss our coarse-grained Langev in dynamics simulations guided by the model that are aimed at designing nanopores for characterizing synthetic polymers. These two examples demonstrate a rational approach of applying modeling and simulation techniques targeted at addressing certain biologically relevant questions.