Enabling Simulation of High-Dimensional Micro-Macro Biophysical Models
  through Hybrid CPU and Multi-GPU Parallelism

There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

Abstract

Micro-macro models provide a powerful tool to study the relationship between microscale mechanisms and emergent macroscopic behavior. However, the detailed microscopic modeling may require tracking and evolving a high-dimensional configuration space at high computational cost. In this work, we present a parallel algorithm for simulation a high-dimensional micro-macro model of a gliding motility assay. We utilize a holistic approach aligning the data residency and simulation scales with the hybrid CPU and multi-GPU hardware. With a combination of algorithmic modifications, GPU optimizations, and scaling to multiple GPUs, we achieve speedup factors of up to 27 over our previous hybrid CPU-GPU implementation and up to 540 over our single-threaded implementation. This approach enables micro-macro simulations of higher complexity and resolution than would otherwise be feasible.

Related collections

Most cited references 13

Record: found
Abstract: found
Article: not found

Large-scale vortex lattice emerging from collectively moving microtubules.

Yutaka Sumino, Ken H. Nagai, Yuji Shitaka … (2012)

Spontaneous collective motion, as in some flocks of bird and schools of fish, is an example of an emergent phenomenon. Such phenomena are at present of great interest and physicists have put forward a number of theoretical results that so far lack experimental verification. In animal behaviour studies, large-scale data collection is now technologically possible, but data are still scarce and arise from observations rather than controlled experiments. Multicellular biological systems, such as bacterial colonies or tissues, allow more control, but may have many hidden variables and interactions, hindering proper tests of theoretical ideas. However, in systems on the subcellular scale such tests may be possible, particularly in in vitro experiments with only few purified components. Motility assays, in which protein filaments are driven by molecular motors grafted to a substrate in the presence of ATP, can show collective motion for high densities of motors and attached filaments. This was demonstrated recently for the actomyosin system, but a complete understanding of the mechanisms at work is still lacking. Here we report experiments in which microtubules are propelled by surface-bound dyneins. In this system it is possible to study the local interaction: we find that colliding microtubules align with each other with high probability. At high densities, this alignment results in self-organization of the microtubules, which are on average 15 µm long, into vortices with diameters of around 400 µm. Inside the vortices, the microtubules circulate both clockwise and anticlockwise. On longer timescales, the vortices form a lattice structure. The emergence of these structures, as verified by a mathematical model, is the result of the smooth, reptation-like motion of single microtubules in combination with local interactions (the nematic alignment due to collisions)--there is no need for long-range interactions. Apart from its potential relevance to cortical arrays in plant cells and other biological situations, our study provides evidence for the existence of previously unsuspected universality classes of collective motion phenomena.