[gmx-users] Force Field for Vacuum simulation
Peter C. Lai
pcl at uab.edu
Thu Feb 2 13:04:50 CET 2012
I don't know if using small molecules in place of lipid chains will save
or use more computer time. On one hand, bond constraints applied to the
lipids will reduce the calculation with respect to degrees of freedom
compared to the same volume of interacting smaller non-polar molecules.
Then again, equilibration time is probably reduced for a large number of
small molecules than for rotamer and translational search of long chains...
so I guess using small hydrophobic molecules might be faster ultimately.
But you may be interested in Schlegel et al 2005 J Mol Model 12: 49-64
Molecular dynamics simulations of bovine rhodopsin: influence of protonation
states and different membrane-mimicking environments for additional
insight.
On 2012-02-02 02:37:53PM +0300, James Starlight wrote:
> Peter,
>
> firstly- thanks for the so detailed discussion.
>
> The negative aspects of water-vacum-water sandwich system wich you provided
> indeed keep me from the modelling of such system.
>
> Indeed I found some information about influence of the specific lipids on
> the stability of the protein by the regulation of tight packing contacts. I
> think that specific cholesterol binding sites in the receptor are very good
> example of such regulation due to heigh regidy of the dual apolar rings of
> that lipid.
>
> How do you think is there any way to simulate such membrane packing forces
> by introducing of the lipid-like layer ? What are the most appropriate
> might be in that case? E.g I've found in the Justin's tutorial cyclohexan
> layer as the membrane-like solvent but I suppose that ussage of the smaller
> apolar mollecule might be more similar to acyl-like interior.
>
> James
>
> 2012/2/2 Peter C. Lai <pcl at uab.edu>
>
> > On 2012-02-02 11:57:22AM +0300, James Starlight wrote:
> > > Peter,
> > >
> > > Yes the main reason is the CPU economy for such experiment. My current
> > > experiment consist of investigation of the tight paking forces (primarily
> > > vdv effect) in the membrane receptor by inclusion of some point
> > mutations.
> > >
> > >
> > > Actyally I think that simple biphastic system ( like in the Justin
> > tutorial
> > > but with vacum layer instead of cyclohexane) is exactly that I need. How
> > I
> > > could make such layers system where water dont move into the middle
> > layer?
> > >
> > >
> > > James
> >
> > The main issue is that vacuum is not a phase. Vacuum is simply empty space.
> > A particle's interaciton with vacuum is null, by definition so a
> > "biphasic" system with one of the "phases" being nothing is literally not
> > comparable to a true biphasic system. You have to realize this from a
> > physical point of view... Moreover, how do you know for certain that the
> > hydrophobic interaction between TM and lipid does not contribute to the
> > TM bundle packing and the interactions you are trying to study? Remember,
> > it has been shown that TMs "may flex to satisfy hydrophobic mismatch".
> > (Yeagle, et al Biochim Biophys Acta. 2007 Mar;1768(3):530-7)
> >
> > Is using united-atom bilayer (Berger lipids, from Justin's KALP-15/DPPC
> > tutorial) still computationally unacceptable for you? (Tieleman et al 2006
> > J. Phys.: Condens. Matter 18 S1221). I did mention earlier also adding some
> > coupling parameters to use a coarse-grained bilayer with all-atom protein
> > too.
> >
> > Anyway, if you are really adamant about not using explicit bilayer,
> > I thought about this a little. Maybe try using explicit waters in the non-
> > membrane portions of the system, then use implicit solvent (GBSA) with
> > the correct dielectric to mimic a hydrophobic phase....
> > I have no methodology for this approach.
> >
> > If I were to take your request of "How do I use explicit water but vacuum"
> > literally: An all-atom system that preserves the vacuum in the membrane
> > region
> > of the system would involve making a 2 walls of immobile waters:
> > You select a monolayer of waters at each end that you want to act as the
> > wall. Rename them to a different residue name in your .gro file. Reorder
> > the .gro file so that all the wall-water atoms are in their own contiguous
> > section. Copy the .itp you use for the mobile waters to another .itp, edit
> > and change the moleculetype to your wall residue name. Add the new .itp
> > to your .top. Either use freeze_grps or position restraints on the wall
> > oxygens to prevent them from moving. This approach has many many problems
> > associated with it, not least that of: "what do you do with the waters
> > in the solvent accessible space" and the fact that since there is vaccum
> > surrounding the outside of the TM bundle and waters inside the TM bundle,
> > your protein may explode from the lateral motion of the waters inside the
> > solvent accessible spaces pushing against the TMs...
> >
> > >
> > >
> > >
> > >
> > >
> > > 2012/1/30 Peter C. Lai <pcl at uab.edu>
> > >
> > > > I am not sure of the actual interpretive value of such a methodology.
> > > > Are you just trying to save computational time by not having to
> > simulate
> > > > an all atom bilayer? The solvent layer is going to contribute to the
> > > > majority of the computation cost in the first place. As an example, the
> > > > bilayer we are using for GPCR work consisting of 238 POPC molecules
> > adds
> > > > 30552 atoms out of a total of 99547 atoms, In any case, people have
> > been
> > > > conducting all-atom simulations of opsins since at least 2005 because
> > > > of the availability of computational resources. See: Lemaitre V.,
> > Yeagle,
> > > > P.,
> > > > Watts, A. Biochemistry 2005, 44, 12667-12680 . So it is unlikely that
> > > > any reviewers today will accept an opsin system simulated in anything
> > but.
> > > >
> > > > Perhaps a biphasic implicit solvent model would work better for you
> > (unsure
> > > > if gromacs can support an "exterior" dielectric and an interior
> > > > dielectric).
> > > > There may also be some people using a coarse-grained bilayer with an
> > > > all-atom
> > > > protein, but again, unless you have carefully determined the coupling
> > > > parameters between a MARTINI bilayer and an all-atom protein, there
> > will
> > > > be doubts about the usefulness of such a system.
> > > >
> > > > If you really must do without the bilayer, you might be able to get
> > away
> > > > with using strong i,i+4 distance constraints within your TMs to
> > preserve
> > > > helical stability while the entire protein is solvated.
> > > >
> > > >
> > > > On 2012-01-30 10:50:35AM +0400, James Starlight wrote:
> > > > > David, Justin
> > > > >
> > > > > Thanks again for help!
> > > > >
> > > > >
> > > > > I have just few questions about in vacum sumulation of membrane
> > proteins.
> > > > >
> > > > >
> > > > > I want to simulate GPCR receptor wich have big transmembrane ( helix)
> > > > > domain and some connecting loop regions wich are exposed to the
> > solvent.
> > > > >
> > > > > As I understood in classical vacum simulation all charges must be
> > redused
> > > > > to avoid its collapse.
> > > > >
> > > > > I want to build biphastic system water-vacum-water where loops would
> > be
> > > > in
> > > > > water and TM helixes in vacum region.
> > > > >
> > > > > Something like this I've read in old publications about simulation of
> > > > > bacteriorhodopsin http://ukpmc.ac.uk/abstract/MED/10412722
> > > > >
> > > > >
> > > > > 1- Could you tell me where I could found possible algorithm about
> > builing
> > > > > of such water-vacum-water system?
> > > > >
> > > > > 2- Also I'd like to specify what should I do with the charges
> > residues.
> > > > I'd
> > > > > like to use AMBER-like or OPLS ff for such simulation As I
> > understood I
> > > > > must neitralize only charges in TM helices and keep residues in LOOP
> > > > > intact. Might this aproache be correct?
> > > > >
> > > > >
> > > > > James
> > > > >
> > > > > 2012/1/29 David van der Spoel <spoel at xray.bmc.uu.se>
> > > > >
> > > > > > On 2012-01-29 17:09, James Starlight wrote:
> > > > > >
> > > > > >> Hi, Justin.
> > > > > >>
> > > > > >> Yes. The GFP chromophore is a part of backbone. It's formed from
> > Ser
> > > > Tyr
> > > > > >> Gly by cyclisation of the Ser with Gly and subsequent
> > dehydrotation.
> > > > As
> > > > > >> the consequence the mature chromophore has cyclised structure wich
> > > > named
> > > > > >> as the CRO residue in PDB structure.
> > > > > >>
> > > > > >> I've made for this CRO residue topology via PRODG server for
> > GROMOS
> > > > ff.
> > > > > >>
> > > > > >> Than I've imported that new chromophore.top into the topology.top
> > of
> > > > my
> > > > > >> structure in accordance to your tutorial.
> > > > > >>
> > > > > >> Finally I've merged CRO.gro and protein.gro ( I've cut CRO from
> > the
> > > > pdb
> > > > > >> for creation of the topoogy for my protein via pdb2gmx)
> > > > > >>
> > > > > >> Than I've done minimisation and chromophore have been diffused
> > from my
> > > > > >> protein :) It seems that I must add covalent bond between CRO and
> > > > > >> protein into the topology. But how I could do it for my multi
> > topology
> > > > > >> file ?
> > > > > >>
> > > > > >
> > > > > > You need to add the bonds angles etc. The easiest way would be to
> > make
> > > > a
> > > > > > special bond (specbond.dat file). You need an rtp entry for your
> > cro
> > > > group
> > > > > > as well. Then pdb2gmx makes the necessary bonds.
> > > > > >
> > > > > > Of course you can make all the bonds, angles, diehdrals and pairs
> > > > manually
> > > > > > as well, but that is tedious and error prone.
> > > > > >
> > > > > >>
> > > > > >> James
> > > > > >>
> > > > > >> 2012/1/29 Justin A. Lemkul <jalemkul at vt.edu <mailto:
> > jalemkul at vt.edu>>
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >> James Starlight wrote:
> > > > > >>
> > > > > >> Hi David!
> > > > > >>
> > > > > >> Thanks for reference I'll study it carefully.
> > > > > >>
> > > > > >>
> > > > > >> I have some general question about the vacuum simulation
> > > > > >>
> > > > > >> 1- I've found that common GROMOS fields are not suitable
> > for
> > > > the
> > > > > >> vacuum simulation because of its implementation for
> > condensed
> > > > > >> phase .
> > > > > >> But In some referencces I've found that people use 53.6 ff
> > for
> > > > > >> the in
> > > > > >> vacuum simulation. In addition Ive done minimisation and
> > > > > >> equilibration
> > > > > >> in that ff in vacuum and my system have not been collapse
> > :) Is
> > > > > >> there
> > > > > >> any wy to adopt this ff for the in vacum ?
> > > > > >>
> > > > > >> 2- I have uncommon onject for simulation. It's GFP protein
> > > > where
> > > > > >> chromophore group ( like ligand) is covalent bonded to the
> > > > > >> backbone of
> > > > > >> this protein. As I've understood in Justins tutorial there
> > are
> > > > > >> no any
> > > > > >> covalent bonds between protein and ligand. But how I could
> > make
> > > > > >> this
> > > > > >> bond if I operate with TWO topology files ( one for
> > > > chromophore and
> > > > > >> another for protein itself) ? I've done all steps in
> > > > accordance to
> > > > > >> Justins tutorial but on the minimisation step my chromphore
> > > > dissuse
> > > > > >> out of the protein interior because of absent of backbone
> > > > group.
> > > > > >>
> > > > > >>
> > > > > >> The GFP chromophore is part of the backbone of the protein, is
> > it
> > > > not?
> > > > > >>
> > > > > >> The tutorial I have for a protein-ligand complex should not be
> > > > taken
> > > > > >> to mean that all non-protein entities are physically separate
> > > > > >> entities. Plenty of cofactors, chromophores, and the like are
> > > > > >> covalently attached to the protein.
> > > > > >>
> > > > > >> -Justin
> > > > > >>
> > > > > >> --
> > > > > >> ==============================**__==========
> > > > > >>
> > > > > >>
> > > > > >> Justin A. Lemkul
> > > > > >> Ph.D. Candidate
> > > > > >> ICTAS Doctoral Scholar
> > > > > >> MILES-IGERT Trainee
> > > > > >> Department of Biochemistry
> > > > > >> Virginia Tech
> > > > > >> Blacksburg, VA
> > > > > >> jalemkul[at]vt.edu <http://vt.edu> | (540) 231-9080
> > > > > >> http://www.bevanlab.biochem.__**vt.edu/Pages/Personal/justin<
> > > > http://vt.edu/Pages/Personal/justin>
> > > > > >> <http://www.bevanlab.biochem.**vt.edu/Pages/Personal/justin<
> > > > http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin>
> > > > > >> >
> > > > > >>
> > > > > >> ==============================**__==========
> > > > > >>
> > > > > >>
> > > > > >> --
> > > > > >> gmx-users mailing list gmx-users at gromacs.org
> > > > > >> <mailto:gmx-users at gromacs.org>
> > > > > >> http://lists.gromacs.org/__**mailman/listinfo/gmx-users<
> > > > http://lists.gromacs.org/__mailman/listinfo/gmx-users>
> > > > > >>
> > > > > >> <http://lists.gromacs.org/**mailman/listinfo/gmx-users<
> > > > http://lists.gromacs.org/mailman/listinfo/gmx-users>
> > > > > >> >
> > > > > >> Please search the archive at
> > > > > >> http://www.gromacs.org/__**Support/Mailing_Lists/Search<
> > > > http://www.gromacs.org/__Support/Mailing_Lists/Search>
> > > > > >>
> > > > > >> <http://www.gromacs.org/**Support/Mailing_Lists/Search<
> > > > http://www.gromacs.org/Support/Mailing_Lists/Search>>
> > > > > >> before posting!
> > > > > >> Please don't post (un)subscribe requests to the list. Use the
> > www
> > > > > >> interface or send it to gmx-users-request at gromacs.org
> > > > > >> <mailto:gmx-users-request@**gromacs.org<
> > > > gmx-users-request at gromacs.org>
> > > > > >> >.
> > > > > >> Can't post? Read
> > http://www.gromacs.org/__**Support/Mailing_Lists<
> > > > http://www.gromacs.org/__Support/Mailing_Lists>
> > > > > >> <http://www.gromacs.org/**Support/Mailing_Lists<
> > > > http://www.gromacs.org/Support/Mailing_Lists>
> > > > > >> >
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >>
> > > > > >
> > > > > > --
> > > > > > David van der Spoel, Ph.D., Professor of Biology
> > > > > > Dept. of Cell & Molec. Biol., Uppsala University.
> > > > > > Box 596, 75124 Uppsala, Sweden. Phone: +46184714205.
> > > > > > spoel at xray.bmc.uu.se http://folding.bmc.uu.se
> > > > > > --
> > > > > > gmx-users mailing list gmx-users at gromacs.org
> > > > > > http://lists.gromacs.org/**mailman/listinfo/gmx-users<
> > > > http://lists.gromacs.org/mailman/listinfo/gmx-users>
> > > > > > Please search the archive at http://www.gromacs.org/**
> > > > > > Support/Mailing_Lists/Search<
> > > > http://www.gromacs.org/Support/Mailing_Lists/Search>before posting!
> > > > > > Please don't post (un)subscribe requests to the list. Use the www
> > > > > > interface or send it to gmx-users-request at gromacs.org.
> > > > > > Can't post? Read http://www.gromacs.org/**Support/Mailing_Lists<
> > > > http://www.gromacs.org/Support/Mailing_Lists>
> > > > > >
> > > >
> > > > > --
> > > > > gmx-users mailing list gmx-users at gromacs.org
> > > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > > Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > > > > Please don't post (un)subscribe requests to the list. Use the
> > > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > > Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> > > >
> > > > --
> > > > ==================================================================
> > > > Peter C. Lai | University of Alabama-Birmingham
> > > > Programmer/Analyst | KAUL 752A
> > > > Genetics, Div. of Research | 705 South 20th Street
> > > > pcl at uab.edu | Birmingham AL 35294-4461
> > > > (205) 690-0808 |
> > > > ==================================================================
> > > >
> > > > --
> > > > gmx-users mailing list gmx-users at gromacs.org
> > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > > > Please don't post (un)subscribe requests to the list. Use the
> > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> >
> > > --
> > > gmx-users mailing list gmx-users at gromacs.org
> > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > > Please don't post (un)subscribe requests to the list. Use the
> > > www interface or send it to gmx-users-request at gromacs.org.
> > > Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> >
> > --
> > ==================================================================
> > Peter C. Lai | University of Alabama-Birmingham
> > Programmer/Analyst | KAUL 752A
> > Genetics, Div. of Research | 705 South 20th Street
> > pcl at uab.edu | Birmingham AL 35294-4461
> > (205) 690-0808 |
> > ==================================================================
> >
> > --
> > gmx-users mailing list gmx-users at gromacs.org
> > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > Please don't post (un)subscribe requests to the list. Use the
> > www interface or send it to gmx-users-request at gromacs.org.
> > Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> --
> gmx-users mailing list gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
--
==================================================================
Peter C. Lai | University of Alabama-Birmingham
Programmer/Analyst | KAUL 752A
Genetics, Div. of Research | 705 South 20th Street
pcl at uab.edu | Birmingham AL 35294-4461
(205) 690-0808 |
==================================================================
More information about the gromacs.org_gmx-users
mailing list