[gmx-users] Error estimation with TI

[Maik Goette]

Thank you David

I am aware of the fact, that FEP seems to be the better way.
I just would like to compare both methods for my system.
Nevertheless another question concerning FEP:
The values in dgdl.xvg have to be treated somehow, I suppose. ;)
I am not sure, if those values are "real" dG- or still dG/dl-values.
Can you tell me, how to treat those values properly, to get the correct 
dG for each lambda run?
And how to treat the dGs from the different lambda runs to get one 
overall dG?

Regards

Maik Goette, Dipl. Biol.
Max Planck Institute for Biophysical Chemistry
Theoretical & computational biophysics department
Am Fassberg 11
37077 Goettingen
Germany
Tel.  : ++49 551 201 2310
Fax   : ++49 551 201 2302
Email : mgoette[at]mpi-bpc.mpg.de
         mgoette2[at]gwdg.de
WWW   : http://www.mpibpc.gwdg.de/groups/grubmueller/


David van der Spoel wrote:
> Maik Goette wrote:
> 
>> Dear all, dear David
>>
>> Again a question concerning error estimation.
>> This time it's due to thermodynamic integration.
>> When I simulate my system, let's say 5 ns, from lambda 0 to 1, I get 
>> my dG/dl values.
>> At the moment, I am simulating both directions (0->1 & 1->0), 
>> integrate over both, and the difference between my back and forth sims 
>> is the error estimate.
>> But I think, one should use a method, similar to FEP, to "guess" the 
>> error from the dg/dl-data, maybe also with generating blocks with 
>> regard to the autocorrelation.
>> But somehow, the delta lambda has to be included, not?
>>
>> Regards
>>
> A much better method is to simulate at e.g. 20 discrete lambda points, 
> check equilibration at each point individually and also get the error at 
> each point. You then finally integrate numerically. Since you check 
> equilibration at each point there is no need to do a reverse sim. as 
> well. Note that it may be good to check the spacing of points, in 
> particular close to 0 and 1.
>