On Monday, March 26, 2012 11:58:43 am Alok Shenoy wrote:
> Hello,
>
> I'm working on a structure using coot, and I have a peculiar problem.
> Searching www didn't quite provide me with an answer, hence requesting more
> knowledgeable users for some help.
>
> Background:
> The structure has 2 chains, Chain A and Chain B. Chain B has all the residues
> that should be present according to our sequence, and the numbering on both
> chains are complete.
>
> Problem:
> 1) Chain A has a couple of residues missing. The numbering shows 231 232 234.
> Clearly marking out the spot where the residue is missing. Sequence shows
> that there is an aspartic acid missing in position 233. Problem is - both 232
> and 234 are connected quite normally, inserting a terminal residue at 232
> (alanine at first) results in absolutely weird connections. Deleting specific
> atoms in the backbone on the protein at that position (between 232 and 234)
> don't quite help either - as that results in 234 missing a peptide bond.
>
> 2) At some points, there is no backbone -by which I mean - lets say residue
> 100 and 101 side chains are showing up (and have appropriate electron
> density). What is missing is the peptide bond that connects these two
> residues (again - electron density for the backbone is present). Is there a
> way to fix this?
>
> Can someone walk me through the solution to this problems?
I would make a copy of the full Chain B by LSQ superposition onto the
oddball Chain A with the fit determined using only residues from the
N-terminus up to the point where the oddness starts.
Then I would walk through the entire copy inspecting the fit to density.
If it looks reasonable, I would throw away the original Chain A and replace
it with the copy of Chain B, followed by rebuilding+refinement as usual.
Ethan
>
>
> Thank You,
>
> Alok Shenoy
>
--
Ethan A Merritt
Biomolecular Structure Center, K-428 Health Sciences Bldg
University of Washington, Seattle 98195-7742
