[gmx-users] Comparing two Ptn-Ptn docking models

[Diogo Martins de Sá]

Hi all,


I have two proteins which I've docked and obtained two models, dock1 and
dock2. Both corroborate, to some point, with literature experiments.


I ran both models in a 76ns simulation, using 43a1 force field and PME.


I have managed to map all salt bridges and their distances through the
simulation (the intermolecular ones! between ARG, LYS, ASP and GLU), all
hydrogen bonds and their percentage of appearance during the simulation
(using one of Justin's scripts), and I also have all other contacts
made between atoms and managed to make index group for all these stuff.


In the .mdp I have specified Protein and Protein2 as energygroups.



I have some ideas but I would like to know what would you guys recommend
I do to infer which of the two models is most likely to be correct. What
useful calculations should I look for? What energy terms (just Short
range ones?)


I guess the best thing to look for would be free energy of binding...
I've been reading about it and came upon David Mobley and Justin's
tutorial, which seem to work using PME (links at the end)
When running the MDS of the ligand (in my case, also a protein) in
water, should I start from scratch? Or can I use tbpconv to extract just
the ligand out of my .tpr and then call mdrun -rerun with the new tpr
and new coupling parameters? Would that satisfy any of the two
tutorials?

thanks in advance
Diogo

Justin tutorial: 
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/free_energy_old/index.html

Mobley's:
http://www.alchemistry.org/wiki/GROMACS_4.6_example:_n-phenylglycinonitrile_binding_to_T4_lysozyme



Thanks in advance,

Diogo