[gmx-users] Position restrain of protein and membrane

[chris.neale at utoronto.ca]

> Thanks Justin for your reply,
> You mean to say that first keep PR on protein allowing the lipids to  
> move(packing), later switch over to production run without keep PR  
> on lipids?
> Here iam getting doubt that, While embedding protein into popc some  
> of the lipids
> will be deleted led to creation of badcontacts, for that require to  
> keep PR on POPC, so it can relieve badcontacts and get minimised  
> structure. am I right?

No. If your lipids are just a 'solvent', then treat them as you would  
treat water. If you are specifically interested in the lipid  
conformations and dynamics then treat them as you would treat water if  
you were interested in the specific solvation and dynamic properties  
of water. Lipid bilayers are essentially a bulk system, the same as  
water, but different from an embedded protein.

Personally, I do 5 ns posre on the protein after insertion and then it  
takes 15+ ns for general protein equilibration. This is from the  
make-hole version of gmx 3.1.4, which I consider to be optimal for  
reducing equilibration time, so your required equilibration time may  
be greater... check time-dependent RMSDs.

On a slightly separate note, I know that many papers do not give a  
full methods description, but many also do. If you have indeed read  
many membrane protein simulation papers and it is your intention to  
question if it is an absolute truth that lipids need not be position  
restrained during equilibration then that is something that you would  
need to elucidate yourself as I have not seen that studied explicitly  
(although I highly doubt that this is the case).

Chris.