David J. Schuller wrote:
> I have an NAD cofactor. Coordinates attached; they were generated by
> running libcheck on the CCP4 dictionary file NAD.cif (standard file
> taken from CCP4 6.1.0). I am trying to rotate various torsional angles.
> I am using the "Edit Chi Angles" feature of Coot 0.6-pre-1. Is this the
> best way? Behaviour seems a bit irregular. I frequently see this message
> (such as when attempting to rotate about var_2 NN1 <--> NC1*):
>
> first atom : NN1: not found in residue
> WARNING:: failure to find all atom pair in residue atoms!
>
> Atom NN1 is in my file (attached). So is every other atom which appears
> in a torsion line with NN1 in the .cif file.
>
> If I choose the top listed bond (var_1 NC7 <--> NC3), the nicotinamide
> ring rotates, but the rest of the dinucleotide doesn't.
>
>
> Attempting var_8 gives a CHI readout, but I can't see anything moving.
> Same for var_9.
>
>
Dear David,
This is a variant of the libcheck/refmac vs Coot NAD bug discussed
previously.
The cause of the problem is the 'bizarre' naming convention of the NN1
and NN7 in the NAD. Coot can't map the dictionary atom name to the name
in the PDB file.
So you have to help out.
You need to quote the NN1/NN7 references in the cif dictionary. IIRC it
should read 'NN1 ' rather than simply NN1 (same for 7).
Paul.
|