Ursula Schulze-Gahmen wrote:
> I am a new user of Coot and I ran into problems with the regularize option
> that I cannot figure out. My model consists of 4 fragments that came out of
> Resolve. I noticed that the 5 C-terminal residues should actually be
> connected between 2 fragments in the middle of the molecule. So, I
> renumbered the C-terminal resiudes so that they would fit into the sequence
> at the correct position. The geometry does not look too bad for connecting
> these residues in their new environment. But if I try to regularize the
> border region between the renumbered residues and the surrounding molecule
> it doesn't work. I think the program still thinks the renumbered residues
> are at the C-terminus.
>
> Error message
>
> click on 2 atoms (in the same molecule)
> (0) N /1/chainid=""/77/LYS, occ: 1 with B-factor: 9.99 element: N at
> (8.724,45.734,24.316) : 0.118271
> (0) N /1/chainid=""/81/ALA, occ: 1 with B-factor: 12.25 element: N at
> (9.137,40.656,20.993) : 0.14991
> regularize: molecule 0 atom index 412 to 2106
> DEBUG:: check_dictionary_for_residues returns OK!
> INFO:: 218 residues selected for refined and flanking residues
> WARNING:: Hit heuristic fencepost! Too many residues to refine
> FYI: 218 > 20 (which is your current maximum).
> Use (set-refine-max-residues 40) to increase limit
> No restraints: regularize()
>
> What do I need to change the place the C-terminal residues in the middle of
> the molecule somewhere?
>
I don't know what is going wrong here. But I can guess at a fix. Either
Extensions -> Modelling Copy Molecule and refine the new copy or
write out the PDB file and read it in again.
If it *still* fails, then you can reorder the PDB by hand before re-reading.
I prefer refinement over regularization.
Paul.
|