Molecular Dynamics Simulation

Question

[Lavanya]

Dear All,

I have modelled a membrane protein and a globular protein and docked them with few natural compounds that we identified. after docking, I thought of simulating both the protein complexes.

1. The membrane protein could not be simulated in desmond, so I am trying discovery studio for it.( Due to time and computational facility limitation, I cannot cary out gromacs)
2. For the globular protein I am currently running dynamics using desmond for 5 ns.

My query is :

1. will a question arise as to why I chose to do simulation using 2 different packages?
2. for both the proteins glide docking with OPLS2005 force field was carried out. but now for simulation of membrane protein, if I use discovery studio that has charmm as the force field, is my methodology wrong? if I carry out simulation in Discovery studio how can I justify it?

kindly suggest

regards lavanya

Answer 1

[vijay]

I think its all about the results that you come up matters the most than the software package that you use. Comparison of packages is fine, but MD simulations are computationally intensive, hence it is better to proceed with single package, but definitely with concrete base data. Dynamics can vary to a certain degree when the force fields changes as their parameters vary. Still you can go ahead with discovery studio, as the basic parameters remains in tact most of the time.

Also ensure that the timescale that you have setup has biological relevance to it (in the sense, does the binding happens in the stipulated timeframe ? else you might stand by the chance of missing out the actual concept that you are studying) .

[Lavanya]

Thanks for the answer. In DS I have run a simulation for 0.5ns for a membrane protein complexed with a ligand..In desmond I have setup a 5 ns simulationf or a 355 amino acid protein complexed with a ligand.. 2 different proteins and 2 diff softwares... yet I am not going to compare the softwares. I chose DS since membrane protein simulation failed in desmond and gromacs.. Is this approach fine to go ahead?

[vijay]

I would rather suggest you to revisit the causes for the failure. Take a look at the solvent models that you have chosen(if any), the time step you have set, the initial and periodic boundary conditions and other parameters of course. Unless you have a strong theoretical data behind the interactions that you are studying, I feel its going to be a cumbersome process, irrespective of the software you use.