|Subject:||Re: [ESPResSo-users] Any possibility to define ion in PeptideB|
|Date:||Wed, 3 Jul 2013 14:41:56 +0430|
The hydrophobocity scale we use for the hydrophobic interactions comes
from a statistical analysis of protein structures. These do contain
charged amino acids and therefore average over electrostatic
contributions (e.g., salt bridges). This is, however, imprinted in the
final hydrophobicity scale in a mean-field way. Otherwise, we
fine-tuned our parameters by focusing on protein alpha3D (PDB: 2A3D).
This protein does contain some charged amino acids.
On Wed, Jul 3, 2013 at 9:59 AM, pdam bright <address@hidden> wrote:
> Dear Tristan,
> Thanks for your complete answer (lots of good information for me).
> One more question! In the current force field, in which the hydrophobicity
> contains electrostatic potential implicitly, had the coarse-graining
> parameters and potential been derived for a neutralized system? (Maybe
> reading the paper you've referred me to, will answer this question.)
> On Wed, Jul 3, 2013 at 11:19 AM, Tristan Bereau <address@hidden>
>> Hi Pee,
>> The model has no neutralizing ions because there's no explicit
>> electrostatics (some of that effect is contained in the hydrophobic
>> interactions of the amino acids, though. See the model's paper [Bereau
>> and Deserno, J. Chem. Phys (2009)] for detail). If you want explicit
>> ions, you'll need explicit electrostatics for the amino acids
>> (otherwise the ions won't do you much good)--so you'd need to
>> reparametrize the force field to take that into account. It's
>> definitely doable, but can be a bit daunting. We had started tinkering
>> with Debye-Hueckel potentials, and there is an early implementation of
>> that in the code. See the $charges==on statement in:
>> At this point we were trying to correct for the hydrophobic
>> interactions when turning on charges for the charged amino acids
>> (because the hydrophobic parametrization contains *some* information
>> about charged interactions). You may want to look into that, though I
>> must warn you we never validated anything with that force field. Once
>> you've parametrized and validated a force field with explicit
>> electrostatics on the amino acids, you could define new beads for the
>> ions--that part of the parametrization shouldn't be too difficult.
>> To define new ions, I'd start looking in src/input/secondary.tcl. This
>> is where the peptide chain gets built (i.e., putting the atoms at the
>> right place, bonding them together). Obviously, building an ion is
>> much simpler! As for the force-field parameters, everything is
>> contained in the files src/chain_*.tcl.
>> To answer your last question: there are obvious shortcomings in not
>> describing electrostatics in a peptide force field. There are a number
>> of scientific problems for which the current implementation is
>> flat-out inappropriate, so in that sense you'd be able to tackle a new
>> class of problems with such a force field.
>> Hope that helps.
>> On Wed, Jul 3, 2013 at 8:28 AM, pdam bright <address@hidden> wrote:
>> > Dear all,
>> > I am simulating a peptide in Espresso using peptodeB. Now I need to
>> > define
>> > neutralizing ions in my system but I found nothing.
>> > My question is that if defining ions is possible, where should I do it?
>> > If not, can I define it by myself? As I've understood, no electrostatic
>> > potential is present in peptideB, So If I can succeed in defining ions,
>> > will
>> > it be significant?
>> > Thanks in advance for your kind help,
>> > Pee
|[Prev in Thread]||Current Thread||[Next in Thread]|