[gmx-users] Regular vs. CHARMM TIP3P water model
nicolas.sapay at cermav.cnrs.fr
Tue Sep 14 13:11:47 CEST 2010
> I don't understand what exactly you want to reproduce.
> Standard tip3p and Charmm tip3p are different models, so the density does
> not have to be identical.
> The Gromacs Charmm FF implementation paper:
> gives 1002 for tip3p and 1015 for charmm tip3p (this is with LJ switched
> from 1 to 1.2 nm).
> I have in an old paper 986 for tip3p, with 0.85 nm cut-off plus dispersion
> These numbers are quite different. I'll do some checking.
Actually, TIP3P and CHARMM TIP3P are supposed to give (almost) identical
results as they are not exactly different models but different
implementations of a same model. That is why I'm surprised of the rather
large difference in the density values observed in my simulations or in
the JCTC paper.
In CHARMM, the additionnal Lennard-Jones parameters for the TIP3P hydrogen
are not here to better reproduce water properties but for practical
purpose - normally, the hydrogens are almost entirely included in the
oxygen VDW radius. As far as I understand, those parameters are here to
improve the stability of simulations by limiting artificial close contacts
between a large negative charge and the point charge on the hydrogen
In fact, I was wondering if the Settle algorithm is appropriate to CHARMM
TIP3P. In the original CHARMM RTF file, there is a SHAKE constraint on the
H-H distance which is not present in the GROMACS tips3p.itp file. So, a
better use of the CHARMM TIP3P model in GROMACS would be to add this
constraint in tips3p.itp and use Shake instead of Settle, isn't it?
>> Date: Tue, 14 Sep 2010 11:00:55 +0200
>> From: spoel at xray.bmc.uu.se
>> To: gmx-users at gromacs.org
>> Subject: Re: [gmx-users] Regular vs. CHARMM TIP3P water model
>> On 2010-09-14 10.23, Nicolas SAPAY wrote:
>> > Hello everybody,
>> > I have many difficulties to reproduce TIP3P simulation results with
>> > TIP3P. Regular TIP3P gives systematically a lower density than its
>> > counterpart, independantly from the cutoff for non-bonded
>> > the version of GROMACS (4.5.1 or 4.0.7) or the double precision.
>> > Regular 961,067 +/-0,756 g/L
>> > CHARMM 980,860 +/-0,492 g/L
>> > The Enthalpy of vaporization follows a similar scheme:
>> > regular -39,992 +/-0,021 kJ/mol
>> > CHARMM -40,665 +/-0,009 kJ/mol
>> How about the dispersion correction?
>> If that is not turned on densities will be too low (in both cases).
>> > In fact, CHARMM gives results closer to what I should obtain at 300K
>> and 1
>> > bar for TIP3P. I suspect an issue with the bond constraints, but I
>> > locate precisely where it is. Settle parameters are exactly the same
>> > well as constraints for flexible water). Did someone ever face a
>> > problem?
>> > Rgards,
>> > Nicolas
>> > P.S. I My system is just a box of 1728 water molecules
>> > for 500 ps at 300 K and 1 bar.
>> > P.S. II I'm using the following simulation parameters:
>> > constraints = hbonds
>> > constraint_algorithm = LINCS
>> > continuation = no
>> > lincs-order = 4
>> > lincs-iter = 1
>> > lincs-warnangle = 30
>> > morse = no
>> > (LINCS or SHAKE should not make any difference since TIP3P is normally
>> > treated by SETTLE).
>> > Other parameters are:
>> > ; COUPLING
>> > ; Temperature coupling
>> > tcoupl = Berendsen
>> > nsttcouple = -1
>> > nh-chain-length = 10
>> > tc_grps = System
>> > tau_t = 0.1
>> > ref_t = 300
>> > ; Pressure coupling
>> > pcoupl = Berendsen
>> > pcoupltype = isotropic
>> > nstpcouple = -1
>> > tau_p = 1
>> > compressibility = 4.5e-5
>> > ref_p = 1
>> > ; NEIGHBORSEARCHING PARAMETERS
>> > ; nblist update frequency
>> > nstlist = 10
>> > ns_type = grid
>> > pbc = xyz
>> > rlist = 1.15
>> > ; OPTIONS FOR ELECTROSTATICS AND VDW
>> > ; Method for doing electrostatics
>> > coulombtype = PME
>> > rcoulomb-switch = 0
>> > rcoulomb = 1.15
>> > vdwtype = Cutoff
>> > rvdw = 1.0
>> > fourierspacing = 0.10
>> > pme_order = 6
>> > ewald_rtol = 1.0e-6
>> > ewald_geometry = 3d
>> > epsilon_surface = 0
>> > optimize_fft = yes
>> > ; RUN CONTROL PARAMETERS
>> > integrator = md
>> > dt = 0.002
>> > nsteps = 200000
>> > comm_mode = Linear
>> > nstcomm = 1
>> > comm_grps = System
>> 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
>> 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
> 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
[ Nicolas Sapay - Post-Doctoral Fellow ]
CERMAV - www.cermav.cnrs.fr
BP53, 38041 Grenoble cedex 9, France
Phone: +33 (0)4 76 03 76 44/53
More information about the gromacs.org_gmx-users