[gmx-users] cyclic peptide: adding constraints

[Mark Abraham]

Giacomo Bastianelli wrote:
> I tried to include the distance contraints in my .top file as follows:

Distance "restraints" are what you want, not "constraints". Read 4.3.4 
in the manual.

Notwithstanding, the stuff below should work so long as your constrained 
atoms are in the same molecule, you've put the [ constraints ] directive 
in the right position with respect to the [ molecule ] directive, the 
molecule is correctly defined, and the starting geometry is not too far 
from the constrained distance.

> [ constraints ]
> ; atom1    atom2    function    distance
> 1    245    2    0.133
> 
> this is my output:
> 
> Fatal error:
> [ file "1NB1.top", line 1665 ]:
>             Atom index (245) in constraints out of bounds (1-1)

grompp thinks the molecule to which you are applying constraints has 
only 1 atom (hence 1-1). Probably you've #included some ion just before 
this, and thus have your [ constraints ] directive associated with the 
wrong [ molecule ] directive, or made some catastrophic error with your 
topology.

> I thought that the function number 2 would have not create
> the bond between the two atoms but just constrained the distance.

Yes.

Mark