[gmx-users] Re:SMD Vs. A and B state

[lloyd riggs]

Dear All,

Sorry, didn't mean to attache the entire huge GMX repost a few minutes back.

In any case, I had questions as follows;

I looked into the following method;

Biophys J. 2008 October 15; 95(8): 3575–3590.
Protein-Protein Interaction Investigated by Steered Molecular Dynamics: The TCR-pMHC Complex, Michel A. Cuendet and Olivier Michielin

I (minus my mistakes) was able to do the same thing by hand by simply EM->NVT->NPT and then a vary short 5 ps or so trajectory for the proteins together, and the same for them apart.

My question is how many runs (seperate traj), in the case of simply a final A and B state,  is necessary to appease critiques that your entire guassian distribution of calculated Free Energy Differences is covered (as in the paper for whole extended pull runs)?  In the paper they do 157+ runs, however it seemed a proof of principle type thing.

I assume a good 10-20 but wanted to ask if anyone had experience with the critique portion.  Basically I wanted to check a good Vs. bad binding affinity, etc...which may be problematic at the ends of free energy curves where a good or bad binding ligand might have overlap ( Vs. the extreems which would be apparenttly different).  I then wanted to do a couple extended pull runs simply to check conformationsal, side chain differences in solvent.

Any advise or suggestions (other than read the tutorial again) is appriciated.  Mostly though I just wanted to get people experience or a rule of thumb.

SIncerely,

Stephan Lloyd Watkins

-- 
NEU: FreePhone - 0ct/min Handyspartarif mit Geld-zurück-Garantie!		
Jetzt informieren: http://www.gmx.net/de/go/freephone