[gmx-users] simulating dimer proteins

[Justin A. Lemkul]

On 6/16/12 7:33 AM, delara aghaie wrote:
> Dear Justin
> Hello :)
> Thanks for your response.
> You have mentioned that "Position restraints are applied per [moleculetype], as
> indicated
> in the topology".
> In my topol.top file there are two .itp files included for the two protein
> chains. I opened the following file:
> topol_Protein_chain_A.itp
> Here I have :
> ---------------
> [ moleculetype ]
> ; Name            nrexcl
> Protein_chain_A     3

Somewhere above this you will see lines something like the following:

#ifdef POSRES
#include "posre.itp"
#endif

This is the information you need.  Both chains should be under the control of 
identical #ifdef constructs, so invoking -DPOSRES will take care of all position 
restraining simultaneously.

> ---------------
> 1)))))) The molecultype has not mentioned explicitly protein as it is required
> by tc_grps (Protein non-Protein). Does gromacs recognizes (Protein_chain_A) as
> protein and so gives special thermostat to it????

Whether or not something is recognized as protein depends on the content of 
residuetypes.dat, so for normal proteins (regardless of how many there are), 
they will be recognized.  Each [moleculetype] is not given its own special 
thermostat, nor should it.

> =====================================
> for the 1RFB protein having two chains, I did energy minimization and then NVT
> simulation for 200 ps and then used g_energy command to plot temperature.
> using this command line:
> g_energy -f nvt.mdp -o temp.xvg    and then select 15   0   for temperature::::
>
> Then it gives me temp.xvg output, but some warning are written on the screen.
> You can see them below:
> ------
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Temperature                 309.834       0.17    3.51504   0.945045  (K)
>
> You may want to use the -driftcorr flag in order to correct
> for spurious drift in the graphs. Note that this is not
> a substitute for proper equilibration and sampling!
>
> 2----((what is the reason for spurious drist and is it necessary to be corrected?))
>

I doubt that warning is relevant.

> WARNING: nmol = 1, this may not be what you want.
> (I do not understand this warning)
>
> Temperature dependent fluctuation properties at T = 309.834.
>
> Heat capacities obtained from fluctuations do *not* include
> quantum corrections. If you want to get a more accurate estimate
> please use the g_dos program.
> 3-----(It has mentioned that this program needs alot of memory. Is it really
> necessary to do this?)
>

None of the preceding messages are relevant, and the fact that they are printed 
is spurious and will be fixed in the code.

> WARNING: Please verify that your simulations are converged and perform
> a block-averaging error analysis (not implemented in g_energy yet)
>
> 4------(What should I do for this warning)???
>

Block averaging is a common analysis used for many observables.  It will tell 
you if you have trends in your data.  For temperature, it should be fairly 
obvious just by looking at the graph whether or not it has converged.

> ========================================================
> 5----Although the two chains of my protein are identical, does it make trouble
> to have two separate .itp files for them? pdb2gmx with the option (chainsep) has
> given these two .itp files to me.
>

It's a perfectly reasonable approach, and in fact the easiest since pdb2gmx 
takes care of everything for you.

-Justin

-- 
========================================

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================