[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Fw: [ccp4bb]: TLS refinement: Fix B-factors?
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
Dear Alex,
Thanks for your reply. I forward the following message to you. Garib says
that the Bfactor fixing to Wilson B does not work yet. Have you checked your
log-file. Since when I fix the Bfactor to Wilson B it is much shorter,
though it does not give any error message. But maybe that only happens to me
since I use ccp4i, I don't know?
Florian
----- Original Message -----
From: "Garib N Murshudov" <garib@ysbl.york.ac.uk>
To: "Florian Schmitzberger" <florian@cryst.bioc.cam.ac.uk>
Sent: Thursday, October 10, 2002 12:20 PM
Subject: Re: [ccp4bb]: TLS refinement: Fix B-factors?
Hi Florian
As I see there is a mistake (possible casued by interface). It uses keyword
whihc is no there. Could you please just use one particular value in FIX B
field. I.e. let say 40 (instead of Wilson B). Program does not understand
Wilson B. Perhaps I should add in future but at the moment it does not work
that way.
Garib
On Thursday 10 October 2002 12:12, you wrote:
> Dear Garib,
>
> Thanks very much for your help. I enclose the two log-files. 3_
refmac5.log
> is the one where I did not fix B-factors in the TLS refinement.
> 4_refmac.log is the one where I followed basically the same refinement
> procedure, but fixing the B-factors to Wilson B in the TLS refinement.
>
>
> Thanks very much for any comments.
>
> Florian
>
>
>
> ----- Original Message -----
> From: "Garib N Murshudov" <garib@ysbl.york.ac.uk>
> To: "Florian Schmitzberger" <florian@cryst.bioc.cam.ac.uk>
> Sent: Thursday, October 10, 2002 11:54 AM
> Subject: Re: [ccp4bb]: TLS refinement: Fix B-factors?
>
>
> Dear Florian,
>
> There is a little bit confusion on interface. TLS parameters box FIX B
> means set initial values to this specific value. In my experinece it does
> not make much difference what you put there. Are you running TLS and
> restrained.
>
> Can you send me two log files so that I could have a look what is going
on.
>
> Regards
> Garib
>
> On Thursday 10 October 2002 11:47, you wrote:
> > Dear Garib,
> >
> > Oh, I think I did not make myself entirely clear. I mean I fixed the
> > B-factors in the TLS parameters-box, not in the scaling Fobs and Fc box.
>
> So
>
> > would you turn the "Fix B-factors" field in the TLS-parameters box on or
> > not? If you turn it on what would you fix the B-factors to?
> >
> > As I understand this B-factor fixing does then only influence the TLS
> > refinement cycles, but not the Refmac cycles. Still I got the impression
> > that Refmac is also doing something differently since the .log-files
> > looks considerably shorter when I turn the "Fix B-factors" in the TLS
> > parameters box. I tried to find where I can specify individual B-factors
> > in ccp4i,
>
> but
>
> > I could not find it. Can I specify it in ccp4i or do I have do run a
> > script?
> >
> >
> > Thanks,
> >
> > Florian
> >
> >
> >
> > ----- Original Message -----
> > From: "Garib N Murshudov" <garib@ysbl.york.ac.uk>
> > To: "Florian Schmitzberger" <florian@cryst.bioc.cam.ac.uk>;
> > <ccp4bb@dl.ac.uk>
> > Sent: Thursday, October 10, 2002 11:17 AM
> > Subject: Re: [ccp4bb]: TLS refinement: Fix B-factors?
> >
> >
> > Refmac does not fix B values. It sets initial value to whatever you give
> > and then refines from that point. If you don't use individual atomic B
> > value refinement then all B values will be equal not necessarily to that
> > given by user. If you use individual B refinement then program will
>
> refine
>
> > TLS and then residual B values. Check if you are using individual B
> > refinement after TLS. That may make difference.
> >
> >
> > Garib
> >
> > On Thursday 10 October 2002 11:09, Florian Schmitzberger wrote:
> > > Dear all,
> > >
> > > I am refining a structure with two protomers in the asymmetric unit;
> > > resolution range 25-2.0 Angstroem; hexagonal space group P6122; total
> > > number of reflections: 22 000; total number of atoms to refine: ~2800.
> > >
> > > I have been using TLS refinement in the last stage of refinement and
> > > got a nice drop in Rfree and Rcryst: before TLS: Rfree=19.3
> > > Rcryst=17.3. after TLS refinement: Rfree=18.2 Rcryst=16.6. I used two
> > > TLS groups for the two protomers in the asymmetic unit respectively, 5
> > > cycles of TLS refinement+9 cycles of normal Refmac refinement. However
> > > I did not turn the "Fix B-factors" field on, though I entered Wilson B
> > > in the
>
> respective
>
> > > field. When going back and runnning the same refinement procedure as
> > > above, this time fixing B-factors to Wilson B I get: Rfree=18.5
Rcryst=
> > > 17.3. i.e little worse than above.
> > >
> > > Since I am not very experienced in using TLS refinement my question is
> > > whether it is necessary to fix the B-factors. It makes more sense to
> > > me, but I would like to have some competent advice on that. If fixing
> > > is advicable: to what should I fix it (with this particular
refinement
> > > problem).
> > >
> > >
> > > Thanks very much,
> > >
> > > Florian
> > >
> > > -------------------------------------
> > > Florian Schmitzberger
> > > University of Cambridge
> > > Department of Biochemistry
> > > 80 Tennis Court Road
> > > CB2 1GA Cambridge
> > > Tel: 0044-1223-766029
--
Dr. Garib N Murshudov, YSBL, Chemistry Dp, Univerisyt of York
Tel: (01904) 328252