[phenixbb] Fwd: Re: Scultor model with gaps included

Gabor Bunkoczi gb360 at cam.ac.uk
Thu May 26 02:44:12 PDT 2016


Apologies, this should have been directed to the mailing list. Gabor

-------- Original Message --------
Subject: Re: [phenixbb] Scultor model with gaps included
Date: 2016-05-25 14:33
 From: Gabor Bunkoczi <gb360 at cam.ac.uk>
To: Joseph Brock <joseph.brock at ki.se>

Dear Joseph,

apologies for the late response, and sorry for the trouble. Please find 
some answers inlined.

On 2016-05-24 15:10, Joseph Brock wrote:
> Dear Phenix community,
> 
> I was hoping if someone could help me with using Phaser correctly,
> since I can't seem to produce the desired search model.
> 
> The first thing that I was wondering about is that the CLUSTAL
> alignment from the log file is quite different to the .aln alignment I
> made with phenix --> other tools --> multiple sequence alignment. Why
> is a sequence alignment required as input if sculptor performs it's
> own internally?

Sculptor can perform sequence alignment, but the algorithm is relatively 
basic, and you can get much better results with an external method (such 
as HHPRED), and hence alignment input is preferred over the target 
sequence. The alignment selected is printed into the logfile for 
reference.

On the other hand, Sculptor does align the structure of your model chain 
to the sequence of your model (this is a surprisingly non-trivial task). 
However, this alignment should be close to 100% identical, unless a 
mistake has been made (such as using the wrong alignment file). You 
should get a warning if the sequence identity is too low. It is normally 
OK to have a couple of mismatches (such as having SeMet in the model 
structure with translates to "X" and not as "C").

> 
> My model sequence aligned to my target sequence in the log file has
> several gaps in the alignment . For e.g.,
> 
> MASTER ......AASFIMV......
> CHAIN_A ......AKM--MF.......
> 
> I would like to great a search model in which the residues aligned to
> gaps in the MASTER chain are included, however the resulting sculptor
> .pdb file has these residues removed. Shouldn't they be inserted so
> that the rest of the chain is in the correct register?

You need to do homology modelling to achieve this, as an insertion can 
affect the whole structure. Sculptor is designed for quick model 
improvement, and just renumbers/renames residues. However, the new 
residues numbers should reflect your sequence, i.e. in your example 
alignment the two methionines on both sides of the gap will get a 
non-consecutive number and are displayed as non-bonded in e.g. Coot.

> 
> If anyone can let me know the parameters to allow this, I would
> greatly appreciate it. I tried unselecting 'gap' in main-chain
> deletion scope and some other variations, but this didn't seem to
> help.
> 
> I seem to get the same result even if I change the target index from 1
> to 2 in the associated .aln alignment. Does anyone know why this is?

Hmm, this does not sound right. Can you send me the logfile so that I 
can have a look (please use my direct email address)?

Best wishes, Gabor

> 
> Apologies if this is quite obvious, I have been reading the
> documentation but can't seem to figure it out.
> 
> All the best,
> 
> Joseph.
> 
> Joseph Brock | PhD
>  Division of Physiological Chemistry II
>  Department of Medical Biochemistry and Biophysics
>  Karolinska Institutet
>  Scheeles väg 2
>  SE-171 77 Stockholm, Sweden
> 
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb
> Unsubscribe: phenixbb-leave at phenix-online.org

-- 
##################################################

      Dr Gabor Bunkoczi

      Cambridge Institute for Medical Research
      Wellcome Trust/MRC Building
      Addenbrooke's Hospital
      Hills Road
      Cambridge CB2 0XY
##################################################



More information about the phenixbb mailing list