AMBER Archive (2004)

Subject: RE: AMBER: Close contact warnings

From: L Jin (s0344557_at_sms.ed.ac.uk)
Date: Thu Feb 19 2004 - 09:11:54 CST

I tried to use ntr=1 to restrain the HO and O connecting to it during
minimizaiton. The result of the minimization seemed good. And the optimized
structure looks good.

 Here is the input file:

 minimization with sander (Amber7)
 &cntrl
  imin = 1, maxcyc = 5000, drms = 0.0005, ntb = 0, ntr = 1, igb = 0,
  ntpr = 50, cut = 999., ncyc = 800,
 &end
Restrain H52
5000.0
FIND
 H52 HO * *
 H53 HO * *
 H54 HO * *
 * O2 * *
SEARCH
RES 1
END
END

                    FINAL RESULTS

   NSTEP ENERGY RMS GMAX NAME NUMBER
   2448 -1.5385E+02 4.8128E-04 1.4965E-03 O36 44

 BOND = 4.8992 ANGLE = 22.6155 DIHED = 18.0337
 VDWAALS = -3.8192 EEL = -431.5714 HBOND = 0.0000
 1-4 VDW = 13.0917 1-4 EEL = 222.7325 RESTRAINT = 0.1644
 EAMBER = -154.0179

But I do not know how this result affects the following MD running with the
command line:

 sander -O -i md.in -o mdout -p prmtop -c min.restrt -r restrt -inf mdinfo -x
mdcrd -e mden

 Here is the input file:

MD in-vacuo, 999.0 angstrom cutoff
 &cntrl
   igb = 0, ntb = 0, ntpr = 100, ntwx = 100, ntwe = 100,

   cut = 999.0, ntc = 2, ntf = 2,

   tempi = 300.0, temp0 = 300.0, ntt = 1, tautp = 0.5,

   imin = 0,
   nstlim = 3000000, dt = 0.001,

 &end

AND THE MDOUT FILES (dt=0.002 has the similar result):

 NSTEP = 0 TIME(PS) = 0.000 TEMP(K) = 457.76 PRESS = 0.0
 Etot = -93.3247 EKtot = 61.4021 EPtot = -154.7267
 BOND = 4.1904 ANGLE = 22.6155 DIHED = 18.0337
 1-4 NB = 13.0917 1-4 EEL = 222.7325 VDWAALS = -3.8192
 EELEC = -431.5714 EHBOND = 0.0000 RESTRAINT = 0.0000
 ------------------------------------------------------------------------------

 vlimit exceeded for step 58; vmax = 21.1388019
 vlimit exceeded for step 59; vmax = 60.4402507
 vlimit exceeded for step 62; vmax = 35.5751264
 vlimit exceeded for step 63; vmax = 512.842258
 vlimit exceeded for step 64; vmax = 26.9734724
 vlimit exceeded for step 65; vmax = 20.0104444

Is there something wrong I have done or I should not do it like this?
Thank you very much.

Lan

Quoting Yong Duan <yduan_at_udel.edu>:

>
> I am not sure assigning vdw parameters without serious rationale is the
> best approach. What happened was that the bond was the one holding H
> from being attracted to O but the H does not have a vdw radii (much like
> in TIP3P water).
> This is not a problem in MD so long as you use rigid bonds for those
> connecting H atoms (with SHAKE), which is required if you intend to use
> dt=2fs. But in MIN, since you did not invoke this, the H can potentially
> get over the small energy barrier. Potential solutions for this, rather
> than playing with the parameters which could have unexpected effect, are
> 1) set small dxm (oddly, dxm will be retired in amber8) or 2) invoke
> SHAKE (and observe that minimization never converge). Sounds horrible
> (particularly for inexperienced), but actually it is really not bad at
> all.
>
> yong
>
> > -----Original Message-----
> > From: owner-amber_at_scripps.edu
> > [mailto:owner-amber_at_scripps.edu] On Behalf Of L Jin
> > Sent: Wednesday, February 18, 2004 1:57 PM
> > To: amber_at_scripps.edu
> > Subject: Re: AMBER: Close contact warnings
> >
> >
> > Thank you Bill. Is it possible to modify some VDW parms to
> > avoid the HO and OS
> > or some other atoms to sit on top of each other? Or what
> > other methods to
> > consider? Thanks.
> >
> > Lan
> >
> >
> >
> >
> > Quoting Bill Ross <ross_at_cgl.ucsf.edu>:
> >
> > > > The type of H54 is HO 'hydroxyl group' within parm99.
> > > > And in my molecule it is attached to the oxygen which is
> > > > part of the sulfate group, -SO3H. I have
> > > > introduced the atom type for the S atom. And O41 is OS in
> > parm99 too. The
> > > > positions of H54 and O41 are like this R-O41-SO3H54. Both
> > of the parms of
> > > H54
> > > > and O41 are from Parm99. Do I need to modify the VDW parm? Thanks
> > >
> > > HO has no vdw. Given that hydroxyls in the Cornell et al. ff
> > > are modeled as vdw-less HO's within the radius of the OH, perhaps
> > > it is worth considering adding a sulfate O type with an even larger
> > > radius, perhaps the same as OW or larger. Here are the radii of the
> > > oxygen types:
> > >
> > > O 1.6612 0.2100 OPLS
> > > O2 1.6612 0.2100 OPLS
> > > OW 1.7683 0.1520 TIP3P water model
> > > OH 1.7210 0.2104 OPLS
> > > OS 1.6837 0.1700 OPLS ether
> > >
> > > Such an approach would keep in the spirit of the Cornell et al. ff;
> > > alternatively one could think of using an H type with a non-0 vdw.
> > >
> > > Bill Ross
> > >
> > --------------------------------------------------------------
> > ---------
> > > The AMBER Mail Reflector
> > > To post, send mail to amber_at_scripps.edu
> > > To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> > >
> >
> >
> > --------------------------------------------------------------
> > ---------
> > The AMBER Mail Reflector
> > To post, send mail to amber_at_scripps.edu
> > To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
> >
>
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu