Re: [ccp4bb] High R factors in Orthorhombic Space Group

- Protein crystallography

Main steps:

- Protein purification
- Crystallisation

Special:

- Programs for crystallography
- X-ray detectors

Basic tutorials:

   - Chemistry
   - Protein
   - Peptide
   - Amino Acids

Xtal community:

- CCP4BB

CCP4bb navigation

CCP4bb <-- 1999 <-- November 1999 <-- 30 November 1999

Previous message:
Subject: High R factors in Orthorhombic Space Group
From: Robert Radford rradford {- at -} UCSD {- dot -} EDU
Date: 2009-12-03

Next message:
Subject: How to add a single water molecule
From: rui ruisher {- at -} GMAIL {- dot -} COM
Date: 2009-12-03

Subject: Re: High R factors in Orthorhombic Space Group
From: mjvanraaij mark {- dot -} vanraaij {- at -} USC {- dot -} ES
Date: 2009-12-03

Rwork = 0.26 and Rfree = 0.32 are not necessarily unacceptably high
for a 2.3 A dataset in my opinion.
Not all crystals are the same, some just have more disorder than
others and not all of this disorder is modelable (if that is a word).
Having said that:
- is the 2.3A data cutoff perhaps a tad optimistic and is your
structure really more like 2.5-2.7A?
- have you correctly modelled solvent/ions etc.?
- did you try TLS analysis and refinement?
- is yours a protein/DNA complex? I understand they often have a
higher R-factors, the specialists may contribute their opinions on the
reason (disorder in DNA? restraints for nucleic acids being less good
than for protein?).
From the information you give, I think you can be confident about the
P21212 spacegroup.
Mark

Mark J. van Raaij
Dpto de Bioquímica, Facultad de Farmacia
Universidad de Santiago
15782 Santiago de Compostela
Spain
http://web.usc.es/~vanraaij/
researcherID: B-3678-2009

On 3 Dec 2009, at 18:49, Robert Radford wrote:

> Hello-
>
> I am working up a data that is giving me unacceptable high R factors.
>
>
> I am solving a data set that is good to ~ 2.3 Å using molecular
> replacement based on a previously crystalized variant. The maps look
> good with all the density fitting well to my model. I originally
> worked up the data using PHASER for the molecular replacement. It
> searched all alternative orthorhombic space groups and found a
> single solution in P2(1)2(1)2 but that only gave me final R factors
> of (Rwork = 0.26 and Rfree = 0.32) Next I went back and tried to
> work the data up in a different space group. Thus far I have tried
> P21 and P1. Both these space groups give me essential the same final
> statistics (~Rwork = 0.26 and Rfree = 0.32). Furthermore, the
> solution looks the same, the only change being the number for
> monomers in the asymmetric unit (1, 2, 4 for P21212, P21 and P1
> respectively) The diagnostics don't indicate that the crystal is
> twined and the data set looks good (i.e. the spots are nice and well
> separated)
>
> The unit cell dimensions for the three cases as
>
> P1: a = 34.55 b = 46.95 c = 88.56, alpha = 89.94, beta = 90.03
> gamma = 89.99
> P21: a = 46.91 b = 34.52 c = 88.47, alpha = 90 , beta = 90.03
> gamma = 90.0
> P21212 a = 88.44 b = 34.51 c = 46.89, alpha = 90.0, beta = 90.00
> gamma = 90.0
>
> One possible complication is that the biological unit is dimeric
> with 2 fold symmetry. The 2 fold symmetry axis for the protein
> appears to be sitting on a 2-fold crystallographic axis.
>
> Any suggestions?
>
> Thank you in advance.
> Robert
> ________________
> Robert Radford
>
> Graduate Researcher
> Dept. Of Chemistry
> UC San Diego

CCP4bb navigation

CCP4bb <-- 1999 <-- November 1999 <-- 30 November 1999

Previous message:
Subject: High R factors in Orthorhombic Space Group
From: Robert Radford rradford {- at -} UCSD {- dot -} EDU
Date: 2009-12-03

Next message:
Subject: How to add a single water molecule
From: rui ruisher {- at -} GMAIL {- dot -} COM
Date: 2009-12-03