[gmx-users] Re: g_lie function
Anton Feenstra
feenstra at few.vu.nl
Wed Jan 24 08:08:11 CET 2007
Oliwia Szklarczyk wrote:
> Dear Mr Feenstra,
>
> I am writing according to Gromacs again. I want to calculate the free
> energy
> of binding of an enzyme to ligand. I am totally new with gromacs. I want to
> use the function g_lie but i don't know how... I read in the internet that
> there should be one input .edr file. But I am supposed to do 2 mdruns for
> the ligand only and for the enzyme-ligand complex right? So I don't
> understand how come there should be one input, or maybe i should somehow
> compile the .edr files from both simulations? I would be grateful for some
> advice, as much as possible for the g_lie function.
g_lie is somewhat badly designed. It will calculate for you alpha*E_VdW
+ beta*E_Coul. But this is only about half of what you need according to
the 'LIE formula' of Aqvist:
delta-G = alpha*(E_VdW,bound - E_VdW,free)
+ beta*(E_Coul,bound - E_Coul,free)
So, what you'd need to do is run g_lie twice, once for the ligand-only
'free' simulation, and once for the enzyme-ligand complex (bound)
simulation. Then you'd subtract the result for 'free' from that of
'bound', yielding the same delta-G estimate as in the LIE formula above.
By the way, the last time I used g_lie, there were some problems with
it. If I recall correctly, one parameter was used as alpha and beta, and
the other was not used. I think this was fixed, but be sure to check!
--
Groetjes,
Anton
_____________ _______________________________________________________
| | |
| _ _ ___,| K. Anton Feenstra |
| / \ / \'| | | IBIVU/Bioinformatics - Vrije Universiteit Amsterdam |
|( | )| | | De Boelelaan 1083 - 1081A HV Amsterdam - Netherlands |
| \_/ \_/ | | | Tel +31 20 59 87783 - Fax +31 20 59 87653 - Room P440 |
| | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/ |
| | "You Think That You're Bigger When You Fill the Void" |
| | (Urban Dance Squad) |
|_____________|_______________________________________________________|
More information about the gromacs.org_gmx-users
mailing list