Hi Jiri,
The distortion during minimization sounds like it could be more
due to model-building artifacts than a basic wrong structure, though
the latter may well also be the case. In such circumstances it would
be interesting to see if proper equilibration could put the duplex in
a local minimum - how long might it last?
Bill
> Hi Bill,
> in this case she likely attempts to simulate a nonexisting
> molecule. The difraction experiment on polyC is old, not much data, and
> oligoC in recent experiments form selfintercalated tetraplexes,
> not duplexes,
> no duplex has ever been seen at atomic resolution.
>
> Jiri
>
>
> > > I'm trying to perform the simulation of double stranded
> > > polyribocytidylic acid but I haven't succeded yet.
> > > I start with the structure built according to the Langridge and Rich
> > > data(Nature,1963): helical rise=3.11 A, helical twist=30 degrees, parralel
> > > strands, each strand consisting of alternating cytosine, protonated
> > > cytosine residues, so that each pair of the double helix contains one
> > > protonated cytosine. But even during the minimization my structure
> > > distorts greatly: bases in pairs become nonplanar and the whole
> > > structure looks rather like two single-stranded helices twisted one on
> > > the another, than a double stranded helix. The more steps of the
> > > minimization I perform, the more distorted structure I obtain.
> >
> > Small, relatively inconsequential errors in the starting structure
> > can induce such problems. Energy minimization is not the best
> > procedure for fixing the model - better would be to use positional
> > restraints on the bases and run dynamics in vacuum, e.g. at 10-100K.
> > (Because of the classic local minimum problem.)
> >
> > It is quite impressive to see a nucleic acid backbone equilibrate
> > under such conditions. After this procedure, solvate, minimize
> > 100-500 steps, and continue with careful equilibration of volume
> > and temperature.
> >
> > Bill
> >
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Sun Feb 19 2006 - 06:10:14 PST
