violation lists & DGII comparisons
Garry W. Buchko
gw_buchko at ccmail.pnl.gov
Wed Sep 18 15:10:20 EST 1996
Because dealing with MSI/BIOSYM is worst than buying a used car from
an ex-O.J.-lawyer turned politician, I have recently been making the
switch-over from DGII to Xplor. Two basic questions.
1) Any general comments about the "quality" of peptide structures
(40-mers, using only NOE based distance restraints) generated by the
two methods and how to best "mimick" DGII conditions with Xplor (DGII:
Tetrangle smooth/4D embed/anneal by heat to 200K and then sigmoidal
cooling to 0K over 10000 step (0.2 psec/step)). I have been
generating Xplor structures with similar convergence by:1)
Full-Embedding, 2) DGSA (2000K for 1000 steps followed by 5000 steps
of cooling to 100K) and 3) Refine: (200K initial and cool in 10000
steps to 1K). The latter refine part makes a substantial difference.
Using "gentle_refine" (200 K and cool for 10000 steps) instead of
"refine" also does a descent job. Is such an overall Xplor approach
reasonable or foolish?
2) NOE violations. The "accept" program lists the NOE violations
(eg: greater than 0.1 A from the distance bounds) for each structure
together. Is there an Xplor program, or does anyone have a program,
that will list all the violations for each NOE restraint for each
structure together? (Eg: for a group of structures, list the
violations for a specific NOE restraint in all the structures
together....similar to the violation list that NMR-Arch generated?).
More information about the X-plor