John Grime received his B.Eng. (hons) in Software Engineering in 2004 from the University of Sheffield, UK, before completing an M.Sc. (Mathematical Biology & Biophysical Chemistry) and Ph.D. (Physical Chemistry) at the University of Warwick (UK). As a graduate student, John studied permeation in biological membranes using microelectrodes and laser scanning confocal microscopy, and through this work he was introduced to computational molecular modelling. John is currently a member of the Voth Group at the University of Chicago, working in both the University of Chicago Dept. of Chemistry and Argonne National Laboratory, where he investigates biological processes using "coarse grained" models and the techniques with which such models can be efficiently applied.
I am interested in how large biological structures such as viral capsids and bacterial carboxysomes self-assemble. These structures are typically composed of large quantities of a small number of proteins, which can automatically organise themselves into much larger structures with important biological functions. Due to the sheer number of interacting particles these processes involve, and the relatively long time scales required (at least on a molecular level!), even the largest current supercomputers cannot perform these calculations using atomic-level models. Instead, I create simplified models which capture the essential physics of the detailed model but require significantly less computer power - particularly if we can remove the individual water molecules from the system to be replaced with an "implicit solvent" approach. These "coarse-grained" models are created using a combination of experimental structural data from techniques such as X-ray crystallography and NMR, along with the results of more detailed simulations. I am also interested in the automatic generation and parameterization of coarse grained models using only experimental structural data alone. Coarse grained representations of biological systems, which can contain large empty regions of space where water molecules would otherwise exist, pose certain technical problems to conventional simulation methods. I am therefore also interested in new techniques which allow these simulations to overcome such problems and make the best possible use of current and future supercomputers. This includes the use of modern graphics cards (GPUs) to increase the rate at which we can perform the simulations and the development of fault-tolerant software to allow simulations to automatically recover from any localized problems on supercomputing hardware.
- Early Stages of the HIV-1 Capsid Protein Lattice Formation, John M. A. Grime and Gregory A. Voth (In review)
- The interactions between charged colloids with rod-like counterions Klemen Bohinc, John M. A. Grime and Leo Lue Soft Matter, 2012, 8 5679-5686
- Free Energy Monte Carlo Simulations on a Distributed Network Luke Czapla, Alexey Siretskiy, John M. A. Grime and Malek O. Khan Lecture Notes in Computer Science, Springer-Verlag, Berlin.
- Decreased Osmotic Pressure via Interfacial Charge Clustering John M. A. Grime and Malek O. Khan Journal of Physical Chemistry B, 2010 114(31) 10049-10056
- The effects of discrete charge clustering in simulations of charged interfaces John M. A. Grime and Malek O. Khan Journal of Chemical Theory and Computation, 2010 6(10): 3205 – 3211
- Interaction between charged surfaces mediated by rod-like counterions: the influence of discrete charge distribution in the solution and on the surfaces John M. A. Grime, Malek O. Khan and Klemen Bohinc Langmuir, 2010 26(9): 6343 – 6349
- Reply to Missner et al.: Timescale for passive transport across bilayer lipid membranes John M. A. Grime, Martin A. Edwards and Patrick R. Unwin PNAS, 2008 105:E124
- Quantitative visualization of passive transport across bilayer lipid membranes John M. A. Grime, Martin A. Edwards, Nicola C. Rudd and Patrick R. Unwin PNAS, 2008 105:14277 - 14282
- Self-assembly and dynamics of the HIV-1 viral capsid protein
- Self-assembly and dynamics of the bacterial carboxysome
- Scalable, fault-tolerant coarse grained molecular dynamics software development
- Automatic CG model generation using "field matching" approaches