Dear AMBER users,
Has anyone of you encountered something similar before? What was the solution?
I run a complex of a protein with a double strand of DNA in water and
compensating salt ions.
The water shell thickness is 5 A (in one case 6); cutoff 9 A; PME and SHAKE are
on. I use split Berendsen temperature coupling with several values for the
coupling parameters (from 0.2 ps to 1 ps), pressure coupling with tau=1 ps,
molecular scaling (npscal=1, and in one case atomic scaling) option. Target
temperature 300 K. The centre of mass motion is periodically removed. I used two
starting structures (X-ray and NMR).
The equilibration was carried similarly as described in the AMBER on-line
tutorial for the DNA molecule.
The problem (or the strange result) is that the side chains rotate very little.
And what is especially disturbing is that even methyl groups do not rotate in a
number of side chains, where one would expect them to rather freely go over
between the three equivalent staggered conformations. The situation improves
somewhat for slower temperature relaxation times (taut=1ps), but even in that
case, only 2 or 3 jumps occur during a nanosecond simulation.
But in that case, the temperature of the solute is fluctuating around 350 K
instead of 300.
The particular methyl groups are on an Ile side chain, which projects into the
major grove of the DNA, and is involved in several contacts with the bases. So
it is in the interface between the molecules and should be able to rotate more
or less freely.
Regards,
Aleks
--
end
-----------------------------------------------------------------------
Aleksandras Gutmanas
Dept. of Biochemistry and Biophysics, tel. +46 31 7787436 (home)
Lundberg laboratory, BOX 462, +46 31 7733947 (laboratory)
Gothenburg University, fax. +46 31 7733910 (laboratory)
S-405 30 Gothenburg e-mail. alex_at_bcbp.gu.se
SWEDEN
Received on Thu May 04 2000 - 08:28:13 PDT
