[gmx-users] Simulation of the protein-ligand system embedded in membrane

[Mark Abraham]

On 26/07/2012 10:14 PM, James Starlight wrote:
> Dear Gromac's users!
>
> I have some questions about simulation of the membrane protein
> complexed with it's ligand in the membrane environment.
>
>
> 1- I want to simulate number of such similar systems ( e.g protein in
> POPC bilayer) which differs only in ligand complexed into protein
> interiour. Should I make my system from the biginning each time ? (
> create protein-ligand complex in vacuu and make minimisation and than
> do solvation in the POPC with further equilibration steps ) On other
> hand if I know coordinates of my ligands in the protein interiour (
> e.g from X-ray structures) is it possible to use apo-form of my
> protein embedded in the membrane and place different ligands in that
> protein-membrane complex? ( on other hand I dont want to do solvation
> each time when I use new ligand )

Only if the apo form has room for you to cut and paste in all your 
ligands... You'll have to equilibrate independently in each case, so you 
don't gain anything by not creating the system from the beginning each time.

> 2- I've found that the ussage of some thermostats could produce some
> artifacts due to the energy distribution of T_coupl algorithms. In my
> current protein-ligand system I want to use  Nose-Hoover's chains
> thermostat. Does it  included in GROMACS ( I've found onlycommon
> Nose-Hoover t_coupl algotithm) and what parametres of that chains
> could be best suited for protein-ligand systems simulated in the npt
> ensemble ?

Have you looked in the manual and read about the options and compared 
the citations it makes to the citations of the algorithm you want to use?

Mark