Hi there all
I am also having problems with the hbdb potential. I have a protein in
complex with AMP.
The PSF is like this:
1 P 1 ALA N NH3 -0.100000 14.0070 0
2 P 1 ALA HT1 HC 0.260000 1.00800 0
3 P 1 ALA HT2 HC 0.260000 1.00800 0
4 P 1 ALA HT3 HC 0.260000 1.00800 0
.......
3926 P 255 LEU OT2 OC -0.570000 15.9990 0
3927 A 300 ATP PG PG 1.20000 30.9740 0
3928 A 300 ATP O1G O1G -0.400000 15.9994 0
3929 A 300 ATP O2G O2G -0.400000 15.9994 0
And when I run the test minimisation script modified for my setup.
Although the hbdb is called it doesn't calculate the H-bonds using the
free format and the hbdb energy is 0. I also tried segm = P and
nothing either.
Any help would be much appreciated. I'm sure its something simple....
Furthermore, has anybody got it working in the fixed Hbond list mode?
Thanks
James
--------------------------------------------------------------------------------------------------------------------------
evaluate ($PARAMETERS2 = "ade.param")
evaluate ($PARAMETERS =
"/usr/local/xplor215/xplor-nih-2.15.0/toppar/parallhdg_new.pro")
evaluate ($STRUCTURE = "LOWESTmin_ATP300.psf" )
evaluate ($TEMPLATE = "mda2_3.pdb" )
param
@TOPPAR:parallhdg_new.pro
@$PARAMETERS2
end
structure
@$STRUCTURE
end
coor @$TEMPLATE
end
!hb database must be read in after psf file
hbdb
kdir = 0.8 !force constant for directional term
klin = 0.08 !force constant for linear term (ca. Nico's hbda)
nseg = 2 ! number of segments that hbdb term is active on
nmin = 1 !range of residues (1st)
nmax = 255 !range of residues (last)
ohcut = 2.60 !cut-off for detection of h-bonds
coh1cut = 100.0 !cut-off for c-o-h angle in 3-10 helix
coh2cut = 100.0 !cut-off for c-o-h angle for everything else
ohncut = 100.0 !cut-off for o-h-n angle
updfrq = 10 !update frequency usually 1000
prnfrq = 10 !print frequency usually 1000
freemode = 1 !mode= 1 free search
fixedmode = 0 !if you want a fixed list, set fixedmode=1, and freemode=0
mfdir = 0 ! flag that drives HB's to the minimum of the directional
potential
mflin = 0 ! flag that drives HB's to the minimum of the linearity potential
kmfd = 10.0 ! corresp force const
kmfl = 10.0 ! corresp force const
renf = 2.30 ! forces all found HB's below 2.3 A
kenf = 30.0 ! corresponding force const
@HBDB:hbdb_files.inp
end
flags exclude * include bonds angles impr hbdb end
mini powell nstep=1000 end
stop
-----------------------------------------
--
James Swarbrick PhD
Lecturer in Medicinal Chemistry
Victorian College of Pharmacy
381 Royal Parade
Parkville
Melbourne
Australia 3052
Tel: ++ 03 99039543
Matthias Buck wrote:
> Dear Charles,
>
> I have been trying to use the HBDB in presence of explicit waters
> (xplor scripting).
> However there are some problems with the assignment/NBonded update
> -
> although I am using different segment names for protein and water
> together with nmin, nmax and nseg (the last just seen in hbdb.inp test
> example) together with segm (just seen under HBDB help),
> I do not believe xplor does a proper selection of the segment.
>
> firstly, there seems to be a limit on the number of residues / atoms: -
> atoms beyond 4729 are not found {and on an energy update
> the HBDB table is not filled}
> -- I don't know what is special about that number, but since the
> complaint
> is about a water which is not in the segment selected, I do not believe
> the selection works.
>
> if I reduce the number of waters in the simulation to get below this
> number, xplor spends a good 2hrs+ on a reasonably fast pc on the
> energy update
> that is meant to give a hbond table - clearly this is not correct.
>
> if I do the HBDB command/assignment before loading in the waters, it does
> much better, now actually identifying the Hbonds and calculating the
> energy,
> but it no longer sees Hbonds in energy updates after waters are added
> (i.e. the psf
> has been modified).
>
> in conclusion I wonder whether there is a bug in the selection HBDB
> passes
> on when there are multiple segments. {has this been tested?}
>
> with best regards,
>
> Matthias
>
> Matthias Buck
> Assistant Professor
> Physiology & Biophysics Department
> Case Western Reserve Univ.
> Medical School, E646
> 2109 Adelbert Road
> Cleveland, Ohio, 44106
>
> fax 216-368-1693
> phone 216-368-8651
>
> e-mail [email protected]
> http://mbalster.cwru.edu/
> _______________________________________________
> Xplor-nih mailing list
> [email protected]
> http://dcb.cit.nih.gov/mailman/listinfo/xplor-nih
-----------------------------------------------------------------------
