Torsion dynamics

0 ehrlich at
Mon Nov 10 05:31:54 EST 1997

Hi all,

I'm trying to use XPLOR for a simulation of Phi/Psi-dynamics via the
'dynamics torsion' feature. I get problems with the setup as XPLOR
terminates with a MAXLEN error which neither Michael Nilges nor I do
understand. I think that I might have gotten the torsion topology setup
wrong. I would be very grateful if anybody of you sees the error I make...
Here's the dynamics torsion statements; a protein with 89 residues has been
read in (segid  's042') and I want keep only the Phi/Psi for the
dynamics. My idea is to fix all torsions first, then free Phi/Psi

============================== tors.xplor
! read in two segments (segid 'a000','s042') before
! fix all torsions first, then iterate over all Psi angles and free them explicitly
dynamics torsion
	maxlen = 5000
		fix group ( all )  ! first keep all torsions rigid
		evaluate ( $ii = 1. )
		while ( $ii < 89 ) loop resnr
			evaluate ( $ii1 = $ii + 1)
			free torsion 	(segid "s042" and resid $ii and name N) 
					(segid "s042" and resid $ii and name CA) 
					(segid "s042" and resid $ii and name C) 
					(segid "s042" and resid $ii1 and name N)  ! Psi
	                ! and now the same for Phi.... not shown
			evaluate ( $ii = $ii + 1 )
		end loop resnr
dynamics torsion
	maxlen = 5000
	reassign = true
	tbath = 300

============================== tors.xplor

XPLOR terminates with the following:
============================== XPLOR's last words

 -------------------------- Torsion Topology ----------------------------------
   WARNING: A chain is too long. 
            Please increase MAXLEN and re-run.
 %TORSION:TOPOLOGY-ERR: Fatal Topology Error
 BOMLEV=    0 reached.  Program execution will be terminated.
 Subroutine DIE called . Terminating

============================== XPLOR's last words

Hmmm. As I set MAXLEN to 5000 already I can't really imagine the chain to
be too long (89 residues should give 178 free torsions if bot Phi/Psi are
considered). Can it be that my way of specifying free torsions (first fix
them all, then explicitly free those you want to simulate) doesn't work as
the same torsion might be kept in the fixed as well as in the free list? 
What would be the 'right' way of running Phi/Psi dynamics with XPLOR?
	Lutz Ehrlich

Lutz Ehrlich
Meyerhofstr. 1
D-69112 Heidelberg

email: lutz.ehrlich at
web  :
phone: +49-6221-387-140
fax  : +49-6221-387-517

More information about the X-plor mailing list