Active Matter Self-Organization Simulator (AMSOS): Combining biophysics and mechanics on HPC
ORAL
Abstract
We have developed a simulation package to study the large-scale behavior of assemblies formed by biofilaments and crosslinkers. Each crosslinker is explicitly tracked in space and its interactions with filaments are described by a two-stage explicit kinetic Monte-Carlo model. Binding-unbinding events of a motor are modeled based on free-energy changes between unbound, singly bound, and doubly bound states, in a way that preserves detailed balance. Crosslinkers binding two filaments walk and diffuse along the filaments thus moving the filaments relative to each other. We use a new numerical method based on geometrically constrained optimization to guarantee that steric interactions and crosslinker binding forces between filaments are properly and efficiently handled. These developments are integrated into a massively parallel software package -- AMSOS -- whose application we demonstrate by a few examples, including aster formation in confinement and self-contracting bundles consisting of more than 100,000 microtubules and motor proteins.
–
Presenters
-
Wen Yan
- Center for Computational Biology, Flatiron Institute, Simons Foundation