Hello Marie-Laurence--
>
> I'm currently trying to calculate the structure of a covalently linked
> homodimer using xplor v2.33 on Ubuntu 10.04 using only distance , talos
> dihedrals, and h-bond restraints. At some point during the calculation,
> there's an error that appears (shown below) about the step size being too
> large then too small. I'm essentially using the same parameters as the
> anneal.py found in the eginput>gb1_rdc example in the xplor package. I have
> also used the same anneal.py file for the monomeric protein and everything
> worked smoothly. The only difference between the monomeric and dimeric
> protein is the number of restraints, which is doubled.
>
> MAKINB: mode 3 found 15831 exclusions, 0 interactions(1-4) and
> 15831 GB exclusions
> MAKINB: mode 3 found 0 exclusions, 0 interactions(1-4) and
> 0 GB exclusions
> *--- Dynamics ---- step= 0 ---- time= 0 ---- delta_t= 0.0015999
> --*
> | E(kin)+E(poten)= 20989.468 E(kin)= 6974.370 temperature=
> 3457.795 |
> | E(poten)= 14015.0982135 grad= 67.5665364 ANGL=
> 2326.9185219 |
> | BOND= 9.1756895 CDIH= 2906.9832523 HBDA=
> 8.1348803 |
> | IMPR= 3231.5519843 RAMA= -0.5304373 VDW=
> 3770.5403615 |
> | noe= 1762.3239610
> |
> *------------------------------------------------------------------------------*
> InternalDynamics::step: large timestep detected. Halving.
Could you please try replacing the RAMA term with the newer torsionDB
term. Setup in your script should be:
from torsionDBPotTools import create_TorsionDBPot
torsionDBPot = create_TorsionDBPot('tDB')
potList.append( torsionDBPot )
rampedParams.append( MultRamp(.002,2,"torsionDBPot.setScale(VALUE)") )
Hopefully, this will solve the problem...
best regards--
Charles
_______________________________________________
Xplor-nih mailing list
Xplor-nih@cake.cit.nih.gov
http://cake.cit.nih.gov/mailman/listinfo/xplor-nih
