School departments and LAS link

New simulation software and high-performance computing permits speedy modeling of complicated biological systems.

Previously, modeling of in vitro biological systems used the assumption that the systems were well mixed and homogeneous. In vivo systems are complex and frequently not homogeneous so these modeling equations do not reflect the true state of these systems. Zaida Luthey-Schulten (Lycan Professor of Chemistry), Elijah Roberts and John Stone have developed the software Lattice Microbes to provide accurate and fast modeling of biological systems using NVIDIA graphics units in high-performance computing.

(a) Model of a crowded E. coli cell with in vivo packing. (b) Sche- matic diagram of the in vivo RDME method.
(a) Model of a crowded E. coli cell with in vivo packing. (b) Schematic diagram of the in vivo RDME method.

Their software incorporates multiple statistical sampling algorithms and can use Message Passing Interface (MPI) to more quickly process the independent trajectories that are generated on large computer clusters. There is ongoing work to increase the processing speed by permitting each trajectory to utilize multiple CPU cores and GPUs. Each trajectory is stored in a HDF5 file and can be used to reconstruct the full distribution of the system's time evolution or to compare with experimentally observed behavior. Various programming languages can be used read the HDF5 files and reaction models contained in SBML files can be imported for initial setup of simulations. For visualization purposes there is a plug-in for Visual Molecular Dynamics (VMD) software that enables VMD to read simulation trajectories for visualization and analysis.

Lattice Microbes was introduced in the January, 2013 Journal of Computational Chemistry. Professor Luthey-Schulten is in the Department of Chemistry at the University of Illinois and a faculty member of the NIH Center for Macromolecular Modeling & Bioinformatics. Co-author Elijah Roberts was a graduate student and postdoc with Dr. Luthey-Schulten and is now an assistant professor at Johns Hopkins. John Stone is an expert on GPU computing and is a Senior Research Programmer at the NIH Center for Macromolecular Modeling & Bioinformatics and until recently was the Associate Director of the NVIDIA Center of Excellence at the University of Illinois.

Studying reactions in biological systems using spatial stochastic simulation is a valuable technique that will be amplified by the use of Lattice Microbes. This software provides high performance exact and approximate methods for sampling the RDME. Both the exact and approximate methods are significantly faster than other tested software. Dr. Luthey-Schulten comments on the use of the software and HPC, "With the availability of high-performance computing (HPC), the method is poised to allow integration of data from structural, single-molecule and biochemical studies into coherent computational models of cells."

A Multi-GPU version of Lattice Microbe that is capable of simulating cellular processes in yeast on the whole cell level over biologically relevant time scales is being implemented on Blue Waters and other NSF/DOE supercomputers by Mike Hallock (SCS Research Programmer) and students of the Luthey-Schulten group.

School of Chemical Sciences
106 Noyes Lab
505 S. Mathews
Urbana, IL  61801
Professor Jonathan Sweedler

(217) 333-5070
(217) 333-3120 fax
jsweedle [at] illinois [dot] edu
Please send comments and suggestions to:
scs-webmaster [at] illinois [dot] edu