I have been using NMR-derived restraints in a simulated annealing protocol to refine the structure of a non-peptidic small molecule. Distance restraints have been working fine, but the torsion restraints have not. In the sander output the torsion restraint is often reported as violated with a value that does not match what I measure with cpptraj.
My molecule has a carbon-carbon double bond in a trans configuration and I have been trying to restrain the torsion between the protons on either side to ~180o. The restraint I am using is
# 1 PAL H18 1 PAL C18 1 PAL C17 1 PAL H17 160 170 190 200
&rst
iat = 53, 3, 4, 34,
r1 = 160.0, r2 = 170.0, r3 = 190.0, r4 = 200.0,
rk2 = 2.0, rk3 = 2.0,
&end
To test the torsion restraint I ran several one step MD runs with just one restraint, the torsion restraint above.
In the starting co-ordinates the value of this torsion as measured by cpptraj was -179.8360o . After a one step MD run the sander output reports the restraint as violated.
------------------------------------------------------------------------------
First atom Last atom curr. value target deviation penalty
------------------------------------------------------------------------------
C18 PLM 1 -- C17 PLM 1: 315.813 190.000 125.813 1.472 t
Total torsion penalty: 1.472
Using cpptraj to measure the torsion after the one step MD gives a value of 179.9927o.
I see the same behaviour with Amber 11, 12 and 14. Have I defined the restraint incorrectly?
Brendan M. Duggan, PhD
NMR Facility Director and Associate Project Scientist
Skaggs School of Pharmacy and Pharmaceutical Sciences
University of California at San Diego
9500 Gilman Drive, MC 0684
La Jolla CA 92093
office phone: (858) 534-8763
lab phone: (858) 822-7826
cell phone: (858) 692-2298
email: bmduggan.ucsd.edu
http://sopnmr.ucsd.edu
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu May 15 2014 - 11:00:02 PDT
