Hi Bert,
- define NCS selections in your input parameters file "ncs.txt" (just
copy-paste-edit whatever is defined automatically, for example);
- do not forget to use "main.ncs=true";
- turn off automatic NCS detection: "ncs.find_automatically=false".
Example:
phenix.refine model.pdb data.mtz ncs.txt main.ncs=true
ncs.find_automatically=false
I'm sure you can do this from phenix.refine GUI as well.
Cheers,
Pavel.
On 3/2/09 5:46 PM, Van Den Berg, Bert wrote:
Hello all,
I'm refining a structure with 4 molecules in the AU. The molecules
have substantial differences in certain regions, so I want to exclude
those regions from the NCS restraints calculation and usage. How do i
do this? As far as I can see, by selecting residues via for example
"chain A and (resseq 20:50 or resseq 88:299), etc" the NCS restraints
are calculated from the specified part of the molecules but still
applied to the entire molecules, which I don't want.
Thanks for any hints!
Bert van den Berg
University of Massachusetts Medical School
Program in Molecular Medicine
Biotech II, 373 Plantation Street, Suite 115
Worcester MA 01605
Phone: 508 856 1201 (office); 508 856 1211 (lab)
e-mail: bert.vandenb...@umassmed.edu
http://www.umassmed.edu/pmm/faculty/vandenberg.cfm
-----Original Message-----
From: CCP4 bulletin board on behalf of Artem Evdokimov
Sent: Fri 2/27/2009 7:00 PM
To: CCP4BB@JISCMAIL.AC.UK
Subject: Re: [ccp4bb] Off topic: Mammalian gene expression in E. coli
Hello,
The short answer is 'yes'. If you can use both methods :) The issue with
limited proteolysis lies in the questionable state of the full-length
protein - if the stuff is nasty and misfolded, then fagments generated by
proteolytic digest aren't going to be meaningful. On the other hand if you
have a small amount of decent quality full-length protein, digest can be
extremely useful. Deuterium exchange is a nice technique if one of your
friends is an altruistic mass-spectroscopist :)
Purely theoretical methods are limited as well, especially if you're
working
with a bunch of unknown domains in a sequence that has low identity with
anything that has associated structures. And in the end, even for known
structures (or very similar ones) truncation can generate surprises - both
positive and negative ones.
Artem
>Hi:
>I am following this with interest. Nice and useful info.
>My question is: how do you "chop the protein into useful hunks"?
>Using some domain identifying software or using limited proteolysis?
>Thanks
>Subbu
