Some afterthoughts:
Of course avoid the common MR problems - assigning wrong SG, saturating
low resolution data etc etc..
1) Any sequence search tool might have told you there was only a poor
match available. 23% is very marginal for MR and with that degree of
similarity you are very wise to start searching for exptl phase methods.
However it is sometimes possible to get an MR solution which can at
least be helpful as a guide while you build your model into the dexptlly
phased map!
BALBES does a very good job at analysing likely models - could your
sequence suggestr a multimer? are there domains which fit well? etc etc.
Then there are the other methods which can give answers - aligning
multiple models, doing homology modelling etc. But even if you found a
reasonable solution rebuilding and refinement can be difficult - much
easier to have some exptl phases to guide that..
Eleanor
Ashley Buckle wrote:
This is generally a good idea, but removal of residues is a subjective process,
and a little trial and error. If your sequence searching finds multiple search
models you can superimpose them and systematically remove the poorer fitting
regions based upon RMSD. We have built a server for this at
http://pxgrid.med.monash.edu.au:8080/mustangserver/
Also see the paper:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0010048
This can generate A LOT of search models, so you need a way of testing them all,
and possibly varying other parameters (eg in PHASER varying RMSD, space groups to
test, clashes). This can generate >100 runs, so best to be able to use a
compute cluster, and some way of queing the jobs (eg MrBump, BALBES etc. Another
solution is here:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0010049
hope that helps
Ashley
On 25/05/2010, at 2:52 AM, Jürgen Bosch wrote:
You've also applied BRAIN 2.0 ?
I mean looked at homologous structures, superimposed them and decided which
parts are to be removed ?
Never trust programs :-) There could be a flexible alpha helix which if you
removed it would have given you in all programs a solution.
it's Monday,
Jürgen
On May 24, 2010, at 10:24 AM, Paul Lindblom wrote:
The last molrep job just finished and it found only an odd solution. So I think
I will try to get my phases elsewhere. But I am somewhat astonished that there
are still enough cases you can't solve by MR.
Thanks to all who replied. Here is a list of servers/programs to find a MR
model:
http://www.ebi.ac.uk/pdbe-srv/view/
http://www.ebi.ac.uk/Tools/fasta33/index.html
http://meta.bioinfo.pl/submit_wizard.pl
XtalPred
http://ffas.burnham.org/XtalPred-cgi/xtal.pl
Balbes http://www.ysbl.york.ac.uk/~fei/balbes/
use the OCA browser for FASTA searches of the PDB
Modeller or Rosetta (both also available as web servers)
ensemble of many proteins with Phaser
FFAS server maintained by the Godzik lab
generate some models using the "Phyre" server (
http://www.sbg.bio.ic.ac.uk/~phyre/ ) and feed the best .pdbs into Mr Bump.
-
Jürgen Bosch
Johns Hopkins Bloomberg School of Public Health
Department of Biochemistry & Molecular Biology
Johns Hopkins Malaria Research Institute
615 North Wolfe Street, W8708
Baltimore, MD 21205
Phone: +1-410-614-4742
Lab: +1-410-614-4894
Fax: +1-410-955-3655
http://web.mac.com/bosch_lab/
Associate Professor Ashley M Buckle
NHMRC Senior Research Fellow
The Department of Biochemistry and Molecular Biology,
Faculty of Medicine
Monash University, Clayton, Vic 3800
Australia
http://www.med.monash.edu.au/biochem/staff/abuckle.html
iChat/AIM: blindcaptaincat
skype: ashley.buckle
Tel: (613) 9902 9313 (office)
Fax : (613) 9905 4699
mobile: 0430 913031
