How you approach this will depend substantially on the sequence identity
between your target and your potential MR models. Definitely remove all
non-protein atoms from your search model, as these are not likely at all
to be present, or present at these positions, in your target. In
addition to using a program like Chainsaw to truncate your model, you
might also consider using models truncated at the N- or C-terminus (if
these are relatively mobile or ill-defined), or consider searching with
multimers (e.g., dimers) for targets that are homo-oligomers with
expected quaternary structure symmetry. For difficult targets in which
there were significant gaps and insertions in the target compared to the
available search models, we used homology-based protein folding
prediction (e.g., Phyre) to prepare search models. For marginal models,
it is possible to get a reasonable crude solution using MR, followed by
density modification and autobuilding to trace a better solution (e.g.
PARROT & BUCCANEER), especially if you have NCS in your target. This
worked really nicely for us for 3UAO, where search models had 30% or
less identity.
Cheers,
_______________________________________
Roger S. Rowlett
Gordon & Dorothy Kline Professor
Department of Chemistry
Colgate University
13 Oak Drive
Hamilton, NY 13346
tel: (315)-228-7245
ofc: (315)-228-7395
fax: (315)-228-7935
email: rrowl...@colgate.edu
On 1/6/2015 8:41 AM, vellieux wrote:
Hello,
Keeping the waters in a model used for molecular replacement search is
not a very good idea.
I'd suggest first that you use a program like Chainsaw and possibly
additional programs to fine-tune your search model to the problem that
you are trying to solve.
Also, reading material such as the ccp4 study weekend proceedings
devoted to molecular replacement and similar material (that are
available on the internet) would be a good idea as well. The problems
you are encountering have been discussed in the literature.
Fred.
On 06/01/15 13:06, allen price wrote:
Dear all:
I got a dataset at 2.8 angstron. I have tried several ways such as
phaser, MRBUMP,BALBES,but still can't solve the
data,which means I have to edit my model. Maybe I'd better cut it off
or delete the water or loop. I really have no
idea,as it is my first time to do such things, I alway used the whole
model to mr. Could anyone give me some advice?
what kind of software do you guys use? really need you help!
Best regards,
Allen
--
Fred. Vellieux (B.Sc., Ph.D., hdr)
IBS / ELMA
Campus EPN
71 avenue des Martyrs
CS 10090
F-38044 Grenoble Cedex 9
Tel: +33 457428605
Fax: +33 476501890
