RE: [AMBER] dftb_disper problem

From: Fernando Martín García <fmgarcia.cbm.uam.es>
Date: Mon, 11 May 2009 13:34:01 +0100

Finally it works. Thank you for your help.

Fernando
El vie, 08-05-2009 a las 08:47 -0700, Ross Walker escribió:
> Hi Fernando,
>
> It is not based on atom type, simply element and the number of neighbors. In your case with H3O+ you may actually be triggering the 4th hybridization state every now and then. The same might be true of the hydrogens, you may end up with 3 atoms close to one of the hydrogens.
>
> Thus you may need to modify the file to read:
>
> O 0.560 0.560 0.560 0.560 3.8 3.8 3.8 3.8 3.15
> H 0.386 0.386 0.386 0.000 3.5 3.5 3.5 3.5 0.8
>
> Of course be aware that this is effectively 'guessing' the parameters here so just keep that in mind when interpreting any results.
>
> Also make sure the code picks up the new file when it runs. The dispersion data should be written to the output file.
>
> Good luck,
> Ross
>
> > -----Original Message-----
> > From: amber-bounces.ambermd.org [mailto:amber-bounces.ambermd.org] On
> > Behalf Of Fernando Martín García
> > Sent: Friday, May 08, 2009 5:43 AM
> > To: AMBER Mailing List
> > Subject: Re: [AMBER] dftb_disper problem
> >
> > Hi Gustavo:
> > I have changed the parameter that you indicated to me, but the problem
> > persist. I was investigating which is the atom that provokes the
> > problems and it's a oxygen atom that belongs to a water molecule, that
> > needs to be protonated (H3O+)at the end of the reaction I'm simulating.
> > These values depend on the type of oxygen atom I'm using?
> >
> > Thanks
> > Fernando
> >
> > El jue, 07-05-2009 a las 09:01 -0400, Gustavo Seabra escribió:
> > > Hi Fernando,
> > >
> > > This is actually a common problem with the use of dispersion in DFTB.
> > > What is happening here is that, at every step, the program tries to
> > > guess a hibridization state for an atom by counting the number of
> > > atoms within a certain distance threshold, in this case, 1.2A. Then,
> > > from the DISPERSION.INP_ONCHSP, it chooses the polarizability data
> > for
> > > this hibridization. The DISPERSION.INP_ONCHSP file has the following
> > > format:
> > >
> > > $ cat DISPERSION.INP_ONCHSP
> > > O 0.560 0.560 0.000 0.000 3.8 3.8 3.8 3.8 3.15
> > > N 1.03 1.03 1.090 1.090 3.8 3.8 3.8 3.8 2.82
> > > C 1.382 1.382 1.382 1.064 3.8 3.8 3.8 3.8 2.5
> > > H 0.386 0.386 0.000 0.000 3.5 3.5 3.5 3.5 0.8
> > > P 1.6 1.6 1.6 1.6 4.7 4.7 4.7 4.7 4.5 !PO4
> > > S 3.0 3.0 3.0 3.0 4.7 4.7 4.7 4.7 4.8 !S not SO2
> > >
> > > The first 4 numbers after the atom name are the polarizabilities for
> > > each hibridization state, defined by the number of neighbors. (To
> > > understand the parameterization, look at the original reference: M.
> > > Elstner, P. Hobza, T. Frauenheim, S. Suhai, and E. Kaxiras, J. Chem.
> > > Phys. vol. 114 (2001) 5149 ) Notice that those parameters have been
> > > derived for DNA bases, so there is no guarantee they will be
> > > appropriate for anything else.
> > >
> > > Basically, these parameters were derived depending on the number of
> > > neighbors you may expect to find for an atom. However, during an MD
> > > calculation, it eventually happens that an extra atom gets close
> > > enough to confuse the program. In your case you have an Oxygen (atom
> > > 2) with 3 neighbors, and there is no polarizability data in the file
> > > for this situation. in this case, sander complains and crashes.
> > >
> > > ============
> > > WORKAROUND
> > > ============
> > > In the Oxygen line, just copy the parameter for 2 neighbors in to the
> > > next column. it will not be precise, but it will preventing sander
> > > from crashing. Line 1 in the DISPERSION.INP_ONCHSP file will will
> > then
> > > look like:
> > > O 0.560 0.560 0.560 0.000 3.8 3.8 3.8 3.8 3.15
> > >
> > > Good luck,
> > > Gustavo Seabra
> > > Postdoctoral Associate
> > > Quantum Theory Project - University of Florida
> > > Gainesville - Florida - USA
> > >
> > >
> > >
> > > 2009/5/7 Fernando Martín García <fmgarcia.cbm.uam.es>:
> > > > Hi all:
> > > >
> > > > I have a problem with the flag "dftb_disper". When i launch my job
> > with
> > > > a value of dftb_disper= 0, the job run well, but if i try to use
> > > > dftb_disper =1, the following message appears:
> > > >
> > > > -------------------------------------------------------------------
> > -------------
> > > > 4. RESULTS
> > > > -------------------------------------------------------------------
> > -------------
> > > >
> > > > *******************************************
> > > > WARNING: a parameter is 0.0 for atom 2
> > > > IZP = 2
> > > > nei = 4 ( 3 neighbors.)
> > > > hh1 = 0.00
> > > > hh2 = 3.80
> > > > Ni = 3.15
> > > > *******************************************
> > > > SANDER BOMB in subroutine <dispersion_params>
> > > > qm2_dftb_dispersion_params.f
> > > > Exiting.
> > > >
> > > > These are the parameters that i use in this job:
> > > >
> > > > &cntrl
> > > > ntx = 7, irest = 1, ntrx = 1,
> > > > ntxo = 1, nmropt = 1, ntr = 0,
> > > > ntpr = 100, ntwx = 100,
> > > > ntf = 1, ntb = 2, dielc = 1.0,
> > > > cut = 9, nsnb = 10, scnb = 2.0, scee = 1.2,
> > > > imin = 0, ibelly=0,
> > > > nstlim = 1000, dt = 0.001,
> > > > temp0 = 300.0, tempi = 300.0,
> > > > ntt = 1, vlimit = 20.0,
> > > > ntp = 1,
> > > > ntc = 1, tol = 0.00001, pres0=1, comp=44.6,
> > > > jfastw=0, nscm=1000,
> > > > ifqnt=1
> > > > /
> > > > &qmmm
> > > > iqmatoms= 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16,
> > 17,
> > > > 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28,
> > > > qmcharge= 0,
> > > > qm_theory='DFTB',
> > > > dftb_disper= 1,
> > > > dftb_maxiter= 70,
> > > > qmshake= 0,
> > > > qmmm_int = 2,
> > > > peptide_corr = 0,
> > > > qmcut= 8.0
> > > >
> > > > I've downloaded the parameter file "mio" from www.dftb.org and
> > installed
> > > > in $AMBERHOME/dat/slko, and also the other two files are in the
> > same
> > > > directory (DISPERSION.INP_ONCHSP and CM3). Can anybody tell me
> > where is
> > > > the problem?
> > > >
> > > > Thanks
> > > >
> > > > --
> > > > Fernando Martín García.
> > > > Centro de Biología Molecular "Severo Ochoa".
> > > > C/ Nicolás Cabrera, 1.
> > > > Campus UAM. Cantoblanco, 28049 Madrid. Spain.
> > > >
> > > >
> > > > _______________________________________________
> > > > AMBER mailing list
> > > > AMBER.ambermd.org
> > > > http://lists.ambermd.org/mailman/listinfo/amber
> > > >
> > >
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > > http://lists.ambermd.org/mailman/listinfo/amber
> > --
> > Fernando Martín García.
> > Centro de Biología Molecular "Severo Ochoa".
> > C/ Nicolás Cabrera, 1.
> > Campus UAM. Cantoblanco, 28049 Madrid. Spain.
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
--
Fernando Martín García.
Centro de Biología Molecular "Severo Ochoa".
C/ Nicolás Cabrera, 1.
Campus UAM. Cantoblanco, 28049 Madrid. Spain.


_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed May 20 2009 - 15:34:32 PDT
Custom Search