[Chimera-users] Chimera Interface to Modeller
meng at cgl.ucsf.edu
Wed Nov 28 10:09:33 PST 2018
The alignment looks fine to me. However, the PDB structures don’t include the N-terminal part of your sequence, so there is nothing to template the model. It is expected that the untemplated portion will simply be added as an extended “string,” as in your image.
Unless you find a structural template for that part, modeling it requires a different method: “de novo” or from scratch. Although there is a choice in Chimera to do untemplated modeling with Modeller (sequence menu: Structure… Modeller (loops/refinement)) it is mainly intended for loops and linkers, not to predict the fold of a whole domain with ~100 residues.
I too ran Blast PDB with your query and did not see any structures that include the first ~100 residues. So I think your choices are to:
(1) delete or ignore this portion in your models, not enough information to model with confidence
- and/or -
(2) look for other programs outside of Chimera to do “de novo” (untemplated) modeling of this domain ( then in Chimera you could stick the 2 pieces together, i.e. with “join models” in the Build Structure tool)
Although I don’t think this applies to your current situation, in general if a sequence alignment seems poor:
(a) You can try re-aligning the sequences using sequence window menu: Edit… Realign Sequences. That will give a choice of program (Clustal Omega or Muscle web service) and other parameters as described here:
(b) You can also edit the sequence alignment manually, but it’s mainly for a minor editing (not too many changes). For a lot of manual editing, you would probably want to use some other program like Jalview.
(c) You can certainly use other software or web servers outside of Chimera to create the alignment as well. As long as you save the alignment in one of the common formats, Chimera can read it from a file.
I hope this helps,
Elaine C. Meng, Ph.D.
UCSF Chimera(X) team
Department of Pharmaceutical Chemistry
University of California, San Francisco
> On Nov 27, 2018, at 5:26 PM, Alisson Oliveira <alisonfeli at hotmail.com> wrote:
> Good night,
> I would like to construct a human DDX3 homology model in its closed conformation. However, I can not perform a proper alignment. I target the sequence Uniprot ID: O00571 and as models the PDB structures: 2I4I and 2DB3. However, the generated models present unsatisfactory results in terms of their quality. Attached is the image of the alignment.
> Can you help me???
> Alisson Felipe de Oliveira
> CRF/RS: 18468
> Mestrando em Ciências Farmacêuticas - UNIPAMPA
More information about the Chimera-users