Working Group Lead: George Karniadakis, gk@dam.brown.edu
Goals and Objectives:
The goal of WG8 is to promote methods in multiscale mathematics and stochastic modeling that can be applied to biological systems. Specific topics include boundary and interface conditions for continuum and atomistic systems, coarse-graining of dense systems, mesoscopic level descriptions, uncertainty quantification, and sensitivity analysis.
This working group is also linked with Working Group 5 High Performance Computing,Computational Issues and Algorithms, and Working Group 9 NanoBiosystems.
Presentations
November 10, 2008 1PM(EST)
Title: Towards Large-Scale Molecular Dynamics Simulations on Graphics Processors
Speakers: Joseph Davis and Michela Taufer (U. Delaware).
Abstract: Molecular dynamics (MD) simulations are a vital tool in chemical research, as they are able to provide a view of chemical systems and processes that is not obtainable through experiment. However, large-scale simulations require access to multicore clusters that are not available to all researchers. Recently many have begun to explore the power of graphics processing units (GPUs) for various general purpose applications, including MD. We present preliminary results of water simulations carried out on GPUs. We also address deficiencies in the implementation of floating-point arithmetic on GPUs and how it affects constant energy MD simulations, which are a way of validating the correctness and stability of an MD algorithm.
Slides Media:08_IMAG_TauferDavis.pdf
Call for Papers
HiCOMB 2009 - Eighth IEEE International Workshop on High Performance Computational Biology
Paper Submission Deadline: November 30, 2008, 11:59pm EST (Extended Deadline)
Held in conjunction with IPDPS-09: 23rd IEEE Intl Parallel and Distributed Processing Symposium, May 25-29, 2009, Rome, Italy
Conference webpage: http://www.hicomb.org/
Journal Articles
Pivkin IV, Karniadakis GE., Accurate coarse-grained modeling of red blood cells. Phys Rev Lett. 2008 Sep 12;101(11):118105. Epub 2008 Sep 12. PMID: 18851338 http://scitation.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PRLTAO000101000011118105000001&idtype=cvips&gifs=yes
Abstract We develop a systematic coarse-graining procedure for modeling red blood cells (RBCs) using arguments based on mean-field theory. The three-dimensional RBC membrane model takes into account the bending energy, in-plane shear energy, and constraints of fixed surface area and fixed enclosed volume. The coarse-graining procedure is general, it can be used for arbitrary level of coarse-graining and does not employ any fitting parameters. The sensitivity of the coarse-grained model is investigated and its behavior is validated against available experimental data and in dissipative particle dynamics (DPD) simulations of RBCs in capillary and shear flows.