[gmx-users] Questions concerning Gromacs

[Mark Abraham]

On 5/01/2011 9:09 PM, Thomas Koller wrote:
> Mark, thanks for your reply.
>
> The only problem are the improper dihedrals. I show you a example:
>
> [ dihedraltypes ]
> ;improper     ftype    phi    k(phi)  mult
> ...
> A  B  C  D     4      180     5.293   2
>
> Sorry, I use ftype 4, not 2 :)
>
> The impropers are listed in the dihedraltypes. The sequence is A B C D where A is the center atom and B, C, D are the surrounding atoms. I think this is right?

When evaluating this function, GROMACS calculates the angle between 
planes ABC and BCD. Whether that angle is what you want to base the 
improper around is something only you know.

> Later the impropers are specified in the [dihedrals]:
>
> [dihedrals]
> ; i   j   k   l    ftype
> ...
>    1   2   3   4      4
>
> I get the error:
> ERROR 1 [file binary_ua.top, line 196]:
>    No default Improper Dih. types
>
> Fatal error:
> unknown function type 20 in ../../../../src/kernel/convparm.c line 326
>
> line 326 is: case F_VSITE2:
>
> Do you see the error? I don't know what is going wrong.

What are the atom types of atoms 1, 2, 3 and 4? Do they match a 
[dihedraltypes] entry in an .itp file you've #included? Order is 
important, but reverses are OK, IIRC.

Mark