Hi All,
I have a general question about the use of residual dipolar couplings
during separate (1) simulated annealing and (2) minimization steps
concerning the initial values for the Cartesian components (alignment
tensor).
I am performing a GB simulated annealing (SA) and minimization using
100 initial structures derived from NMR restraints, using one set of
initial values for the alignment tensor (s11...s23) for all 100
structures. The alignment tensor floats during the SA, and the
coordinates are written at the end of the SA. Next for the
minimization step, the coordinates from the SA step (where the
alignment tensor was allowed to float) and the original dipolar
couplings (initial alignment tensor values) are read in. Because the
alignment tensor was allowed to float during SA, high Align energies
(>100 kcal) are reported for some structures during and at the end of
the last minimization step, when compared to the alignment tensor
information (s11...s23) in the DIPOLE restraint file. Some structures
have a more reasonable Align energy (<10 kcal), but the energies are
on average higher compared to the Align energies at the end of the SA
step (float tensor). To resolve the high Align energies:
(A) Should the alignment tensor information (s11...s23) in the DIPOLE
restraint file be updated for the last minimization step to reflect
the minimized (float) alignment tensor from the SA step?
(B) Should I increase the number of steps of either the steepest
descent or conjugate gradient minimization performed during the last
minimization step to allow the tensor to float back to the original,
initial estimate?
(C) Should I not specify a DIPOLE restraint file in the minimization
step?
Thanks,
Doug
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri Apr 18 2008 - 21:14:15 PDT
