[gmx-users] Size of perturbed groups in FEP
mgoette at mpi-bpc.mpg.de
Wed Jan 11 17:25:11 CET 2006
Shouldn't be a problem, though.
For ILE (and except for proline/glycine any other) "just" turn the
Ca-Proton into the "new" Cb, add the other ILE-atoms as dummies
(structurally correct, though), turn the original Cb into the proton,
and let the original ILE-atoms vansih to dummies.
With softcore, the convergence should be reached....hopefully.
I think, it depends strongly on the environment due to your peptide.
Nevertheless, correct error estimation is really very important, though.
Maik Goette, Dipl. Biol.
Max Planck Institute for Biophysical Chemistry
Theoretical & computational biophysics department
Am Fassberg 11
Tel. : ++49 551 201 2310
Fax : ++49 551 201 2302
Email : mgoette[at]mpi-bpc.mpg.de
WWW : http://www.mpibpc.gwdg.de/groups/grubmueller/
David Mobley wrote:
> I'm not an expert on free energy calculations involving mutation of
> one thing into another; most of the calculations I do involve
> disappearing of things. In that case (that is, the case of, say,
> disappearing an isoleucine sidechain) you can certainly do it. It will
> probably, however, be a lot harder to disappear the side chain in one
> form and then appear it in another; it's hard enough to get converged
> answers for the former, IMO. I think I might think about mutating both
> to a common reference state so that you can get the free energy
> difference by doing two disappearing calculations. Obviously this
> takes "longer" than doing one calculation to bridge the whole gap, but
> on the other hand it may be easier to get good convergence.
> Others who have experience with this may be better prepared to comment.
> One other thing: For these calculations, careful error analysis is
> very important (in order to be sure you're not just seeing noise). You
> need to take into account the time correlation in the data series; the
> standard error isn't simply the square root of the variance divided by
> the number of samples -- you need to correct for the fact that the
> samples are correlated. This can make a big difference as the
> correlation time is often fairly long compared to 1 or 2 fs.
> Best wishes,
> David Mobley
> Postdoc, Dill lab, UCSF
> On 1/11/06, Bruce Milne <bfmilne at ff.up.pt> wrote:
>>I have been wondering about the possibility of doing free energy
>>calculations on D- and L- forms of amino acids incorporated into small
>>peptides but I have no experience in this area and I was just wondering
>>if it is possible to do free energy perturbations on groups as large as
>>the side chains in e.g. isoleucine. It seems to me that changing a
>>hydrogen atom for, say, a methyl group might be OK but what about
>>something as large as -CH(CH3)CH2CH3 ?
>>I guess in practice the unit to be perturbed would have to include the
>>backbone carbon plus the hydrogen and side chain in order to go from the
>>D- to the L-form of an amino acid.
>>Sorry if this is really basic and obvious but any advice would be
>>Dr. Bruce F. Milne PhD
>>Faculdade de Farmácia
>>Universidade do Porto
>>Rua Aníbal Cunha - 164
>>email: bfmilne at ff.up.pt
>>Faculdade de Farmácia: http://sigarra.up.pt/ffup/web_page.inicial
>>Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com
>>gmx-users mailing list
>>gmx-users at gromacs.org
>>Please don't post (un)subscribe requests to the list. Use the
>>www interface or send it to gmx-users-request at gromacs.org.
> gmx-users mailing list
> gmx-users at gromacs.org
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
More information about the gromacs.org_gmx-users