Molecular dynamics (MD) simulations are a powerful scientific tool used to study a wide variety of systems in atomic detail. From a standard protein simulation, to the use of steered molecular dynamics (SMD), to modelling DNA-protein interactions, there are many useful applications. With the advent of massively parallel simulation programs such as NAMD2, the limits of computational analysis are being pushed even further.
Inevitably there comes a time in any molecular modelling scientist's career when the need to simulate an entirely new molecule or ligand arises. The technique of determining new force field parameters to describe these novel system components therefore becomes an invaluable skill. Determining the correct system parameters to use in conjunction with the chosen force field is only one important aspect of the process. The ability to use several programs simultaneously to examine and refine the system in question is also a critical element of these kinds of problems. It should be noted that the extent of parameterization carried out in this exercise is minimal and designed to emphasize the major points required in a more detailed fitting procedure. Road-maps for more systematic optimizations that include experimental data can be found in a series of articles for CHARMM [1,2], AMBER , and other force fields, including OPLS-AA [4,5] and ECEPP. A recent paper by Autenrieth et al. reports both AMBER and CHARMM parameterizations for the reduced and oxidized forms of cytochrome C and is a useful reference for parameterization methodologies . Additional sources for parameterization can also be found on the web (see , , , ). Polarizable force fields that include terms to allow polarization of the charge distribution by environment are under development .
This tutorial will walk you through a comprehensive example of how one investigates, sets up, and simulates a small nonstandard ligand bound to a protein system; specifically, we will investigate the glutaminase subunit of the hisH-hisF system and will determine parameters for its covalently bound substrate. Starting from the crystal structure in the protein database and using the breadth of available biochemical information, we will dock the substrate (glutamine) to the active site of hisH and develop the missing parameters in accordance with the CHARMM force field. As a first guess for the parameters, we will try to derive as many of the missing parameters as possible from existing similar molecules already parameterized in the force field. Then we will further refine these new parameters with semi-empirical quantum chemistry calculations. Once the new parameters are finalized, we will minimize the system. The combination of all of these techniques will require the use of at least three different computational biology and chemistry packages.
The entire tutorial takes about 2-3 hours to complete.
This tutorial assumes that VMD  (version 1.8.3 or higher), Spartan  (optional, but if used, version R04 is preferred), NAMD2  (version 2.5 or higher . . ) and a text editor has been correctly installed on your computer.