[gmx-users] Ligand binding energy using LIE with PME

[Diane Fournier]

On the same subject, I have read in the list about long-range interactions (Coul-LR, LJ-LR) which are supposed to be included in the energy parameters calculated in g_energy, but I never see those. Do you have to write something specific in the .mdp files to get them ?

Also, some of my inhibitors and complexes are charged. In my simulation of the complex, I have added a counterion (Na). Does this complicate matters worse ? For the inhibitors in water, I protonated the ligand, according to a procedure I saw in a paper.

Would it be simpler, in the case of the complexes, to protonate one residue away from the active site ?


-----Original Message-----
From: gmx-users-bounces at gromacs.org on behalf of Diane Fournier
Sent: Wed 8/16/2006 10:10 AM
To: gmx-users at gromacs.org
Subject: [gmx-users] Ligand binding energy using LIE with PME
 
Hi gmx-users

I know that this has been discussed before from consultation of the mail archive, but there is still confusion :

I want to obtain the ligand binding energy from my 1ns PME simulation of a ligand-protein complex. I have made, as directed, a simulation (with PME) of the ligand in water to calculate Eqq and Elj. I am not using the g_lie program since I know I will need to mind the PME contribution.

I'm using (I hope I am correct) the average values of all ligand-protein (Coul-SR, Coul-14, LJ-SR, LJ-14) and ligand-rest (Coul-SR, Coul-14, LJ-SR, LJ-14) terms for the ligand complex simulation and all ligand-solvant terms (same) from the ligand in water simulation. In my case the 1-4 and ligand-rest terms are equal to zero, which I think is normal as no 1-4 interactions are defined between ligand and protein (since they are not covalently bonded) and that no restraints are applied.

It is for the so-called PME long-range contribution that I get confused. I have done a mdrun -rerun of my trajectory using a .tpr file where the ligand has zeroed partial charges for both simulations. From what I understood, the same energy terms as mentionned above have to be extracted from these rerun simulations and substracted from the previous terms as they represent what the ligand "feels" from PME from the image systems around the box. In my case, these contributions are very small (< 1kJ/mol), which I think is due to the fact that I am using a box size (d = 0.85nm for the protein, d = 2.0nm for the ligand) which is supposed to make PME artifacts negligible.  

I am just wondering if this is correct so far, or if I need to do other rerun simulations with zero charge on protein and solvant to obtain the PME contribution.  

Any input will be appreciated

Diane    


-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 4033 bytes
Desc: not available
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20060816/92a866c9/attachment.bin>