Dear David,
thank you for the answer.
>Pick some intense peaks whose assignments are clear, and determine oscale
such
>that the calculated and measured intensities become as close as they can.
Really, oscale was wrong, now it's ok.
I ran NMR restraint calculations with distance restraints but want to learn
volumes restraints calculations too. Also, ain't the calculations with
volumes restraints more straightforward since the experiment provides us
with the volumes, not distances, and there is no need to calculate
restraints by other programs?
Could I ask you some more questions about vol. rest. calcs?
- Since my molecule isn't very big (630 atoms, see attachment), would it be
better to use only one submolecule (entire molecule)?
- The molecule structure changes during the MD run so do the calculated
values of volumes. Should I periodically change oscale to maintain the
calculated and measured intensities at the same level?
- If I keep default value of the NOESY weight (=1) the NOESY energy (~30)
is much less then EAMBER (~-5000), so NOESY restraints seem to have no
influence on the simulation. What value of NMR volumes restraints energy is
appropriate for minimization/MD/annealing at T=300K? Should it be
comparable with the dihedral angles energy, for instance?
- I want to use values of Pearson r, Rms error, R1 and Rr as criteria of
whether the structure is in agreement with experimental data, but there are
not a lot of works with NOESY volumes restraints, so could you tell what
values of these factors are considered as appropriate for nucleic acid
simulations?
P.S There is only one step of minimization in output file. This was done to
see the statistics that show itself only after the calculation on the one
CPU. Previous minimization was performed on the several CPUs.
Respectfully,
Eugene
On 6 July 2015 at 20:42, David A Case <david.case.rutgers.edu> wrote:
> On Mon, Jul 06, 2015, Eugene Sviridov wrote:
> >
> > I'm trying to solve DNA-duplex structure using MD with NOESY volumes as
> > restraints.
> > When I try to minimize the structure it looks like the molecule doesn't
> > feel any restraints: I tried to increase restraints weights, to use
> random
> > volumes values and no volumes restraints at all - resulting structures
> are
> > nearly the same (<0.5 RMSD).
>
> We would need lots more information to be of much help. You should be
> able to
> see from the mdout file whether you have NOESY penalties being applied or
> not.
> In particular, you can see whether or not the calculated and measured
> intensities are close to each other, etc.
>
> There are example files in $AMBERHOME/test/noesy that you can compare to
> what
> you are getting. But, as a beginner, you should certainly run NMR
> restraint
> calculations with "conventional" restraints (as in Tutorial A4) before
> trying
> NOESY volume restraints (which is most definitely a very advanced
> application.)
> >
> > Could it be connected with a wrong value of "oscale" parameter? I failed
> > to find any information about its determination. How could I estimate it?
>
> Pick some intense peaks whose assignments are clear, and determine oscale
> such
> that the calculated and measured intensities become as close as they can.
>
> ....dac
>
>
> _______________________________________________
> 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
- application/octet-stream attachment: ds.prmtop
- application/octet-stream attachment: minh.out
Received on Fri Jul 10 2015 - 04:00:03 PDT
