Re: [ccp4bb] CH-bond length discrepancies

[Santarsiero, Bernard D.]

It totally depends on what you define as your C-H bond length.

For XRD work, the C-H is the distance between the electron
center-of-masses. For  neutron or high-resolution XRD work, you can
determine the distance between the nuclear center-of-masses.  They are
different by about 0.1A.  The electron COM for hydrogen is substantially
shifted along the bond towards the carbon atom.

I did a very precise determination of 1,1-dicyclobutanedicarboxylic acid
at 20K, and the C-H bond lengths averaged to 0.991A, and far shorter than
the 1.096A you'd expect from a neutron diffraction study.

The VDW distances shouldn't be that important since the potential isn't
that sensitive with a difference of 0.1A.  If you have to restrain C-H
distances for XRD refinement, then they should be shortened.

BDS



On Sat, April 29, 2017 8:22 am, benjamin bax wrote:
>
>
>
> 'Most distances   between bonded  atoms   weresettled longago 
> to  highaccuracy,   but,in  the caseof  
> hydrogens,  the values  in  common  use often   differ  by
>   as  muchas  20%.'
>
> Phenix/   MolProbity  HydrogenParameter   Update
>
> Deis, L. N., Verma, V., Videau, L. L., Prisant, M. G., Moriarty, N. W.,
>
> Headd, J. J., Chen, V. B., Adams, P. D., Snoeyink, J., Richardson, J. S. &
> Richardson, D. C. (2013). Comput. Crystallogr. Newsl. 4, 9–10.
>
>
>
> On 28 Apr 2017, at 18:33, Bernhard Rupp  wrote:
>
> Dear Fellows of the Bond,
>
> when validating a QM refined homology model with Molprobity, I noticed
> various 8 sigma deviations in the carbon-hydrogen bond distances.
> Out of curiosity, I then used refmac to calculate riding Hs for the same
> model, and at least in one instance (N-H backbone) there are
> significant differences between Molprobity and Refmac H bond distances
> (differences to the QM distances in other
> instances I find interesting, but less relevant for us).
>
> The riding H vs Molprobity presumably should be consistent, because if we
> use them in VDW restraints but
> they differ from the validation target, systematic bias will occur. I have
> no feel how significant that effect
> might be – maybe someone more erudite can comment.
>
> Examples
>
> distance  MP   REF QM
> backbone N-H   0.861.011.00
> phenyl C-H 0.930.931.09
>
> Best, BR
>
> PS: If someone has accurate experimental values for CH distances I’d
> appreciate a link.
> No access to CSD.
>
> --
> Bernhard Rupp
> Crystallographiae Vindicis Militum Ordo
> http://www.hofkristallamt.org/ 
> b...@hofkristallamt.org 
> +1 925 209 7429
> +43 767 571 0536
> --
> Many plausible ideas vanish
> at the presence of thought
> --
>
>
> Dr Ben Bax
>
> York Structural Biology Laboratory,
> Department of Chemistry,
> University of York,
> York YO10 5DD
>
> ben.d.v@gmail.com
>
>
>
>
>
>


-- 
Bernard D. Santarsiero
Associate Director, UI Health Systems Biorepository
Research Professor
Center for Biomolecular Sciences and the
  Department of Medicinal Chemistry and Pharmacognosy
Center for Structural Biology
Center for Clinical and Translational Science
University of Illinois at Chicago
MC870  3070MBRB  900 South Ashland Avenue
Chicago, IL 60607-7173  USA
(312) 413-0339 (office)
(312) 413-9303 (FAX)
http://www.uic.edu/labs/bds
http://scholar.google.com/citations?user=fGauLBMJ

[ccp4bb] REFMAC - TWIN OPTION

[Santarsiero, Bernard D.]

Can someone point me to bulletpoint documentation on using the twin
refinement in CCP4?

Here's what I did.

1.  I'm in space group P3, and the see a very clean diffraction pattern
that looks like one single lattice. Very clean spots, so merohedral
twinning.

2.  You can use various programs to estimate the twin fraction and select
various twin laws.

3.  You run DETWIN in the UTILITIES section of DATA REDUCTION and
ANALYSIS. I have both intensities IMEAN and amplitudes FMEAN, and selected
amplitudes. The FMEAN changes to something like FMEAN_detw. You have to
select a twin fraction, but does it matter what the twin factor you assign
is?  (It seems to get refined, or just estimated, in refinement?)

4.  There are potentially 3 twin laws, so do you run DETWIN three times? 
You can, in fact, get amplitudes that are FMEAN_detw_detw_detw.  I've run
it once and three times, and seems to give the same result in REFMAC
refinement.

5.  Select the Free R subset. Will uniqueify choose correctly?

6.  In REFMAC, it looks for the F_detw amplitude, or you can select it.
Then it seems to select and test three twin laws, even if you only ran
DETWIN once, with one selected twin law. It tests if the twin laws give
low R(merge).  If not, it tosses them out. It seems like the R(merge)
calculation depends on the starting model, and not just on the modified
Fobs data. It also seems like it tries to calculate the twin fraction, and
then refine the structure. Or, is it just giving an estimate, and I need
to go back to original data and rerun DETWIN?

7.  I tried to work with intensities instead of amplitudes, and it can
blow up. I've also refined, got a good refinement (3 twin laws, each with
twin fraction around 0.1, Rcryst = 19%, Rfree = 21%), rebuilt one chain of
the structure, and it blows up, suggesting no twinning, but Rs around 35%.

Does REFMAC refine the twin fraction?
Do I have to assign a twin fraction in DETWIN, or does it just set up the
relationships between reflections and amplitudes?
Should I try to work with intensities or amplitudes.  The calculation is
presumably done on intensities, but seems to be more unstable using that
option.

I haven't tried the PHENIX option. Just a quick step-by-step guide on what
to do would be useful.

Thanks,

Bernie
-- 
Bernard D. Santarsiero
Research Professor
Center for Structural Biology
University of Illinois at Chicago
MC870  3070MBRB  900 South Ashland Avenue
Chicago, IL 60607-7173  USA
(312) 413-0339 (office)
(312) 413-9303 (FAX)
http://www.uic.edu/labs/bds
http://scholar.google.com/citations?user=fGauLBMJ

[ccp4bb] ATP library file in REFMAC - Part II

[Santarsiero, Bernard D.]

I appreciate all of the comments and suggestions on how to locate a better
CIF file for ATP. I adopted the suggestion of Boaz, and generated a new
ATP.cif file, and refined one of my structures.

The validation server still uses substantially different target bond
lengths and angles, so there is better agreement, but I still get flags.
Both bond length and angles target values are more realistic than what was
used in the REFMAC ATP.cif file, but I still find fault with the target
values.

For example, the bond lengths and angles around PB, the beta P atom, range
from 104.4-108.7deg in the grade file, and are unrealistic since they
are all less than the tetrahedral angle. The CIF also targets one P-O as a
single and the other as formally a double bond, but most crystallographic
studies indicate at least some delocalization over both internal oxygen
atoms, and a slight compression of that angle.  The validation server has
an ideal target value of 101.66deg which is surely contracted from a
more reasonable value near 105deg.

One approach that would be useful is to consider refinement on special
groups, like phosphate, with structural parameters, instead of actual
target bond lengths and angles, so R1 for the two PB-O1B and PB-O2B
distances, and R2 for the two PB-O3A and PB-O3B distances, and pairs of
angle restraints incorporating a complete delocalization model. See, for
example, Murray-Rust, Burgi, and Dunitz, J. Am. Chem. Soc., 97, 921
(1975), or Tamasi at al., The Open Crystallography Journal, 3, 1 (2010).
This could be used to directly address questions about delocalization and
hydrolytic capacity, rather than trying to determine then from bond
lengths.

Bernie




On Wed, August 27, 2014 2:05 pm, Boaz Shaanan wrote:
   Or use Grade:
 http://grade.globalphasing.org/cgi-bin/grade/server.cgi
   which gives the correct bond length.
   Boaz

   Boaz Shaanan, Ph.D.
  Dept. of Life Sciences
  Ben-Gurion University of the Negev
  Beer-Sheva 84105
  Israel

  E-mail: bshaa...@bgu.ac.il
  Phone: 972-8-647-2220  Skype: boaz.shaanan
  Fax:   972-8-647-2992 or 972-8-646-1710

   From: CCP4 bulletin board [CCP4BB@JISCMAIL.AC.UK] on behalf of
 Bernard D Santarsiero [b...@uic.edu]
  Sent: Wednesday, August 27, 2014 6:12 PM
  To: CCP4BB@JISCMAIL.AC.UK
  Subject: [ccp4bb] ATP library file in REFMAC

 I recently refined a structure in CCP4/REFMAC with ATP in the
 structure. Upon submission to Acta for publication, the wwPDB
 validation report was run. Several things were flagged, including the
 C4-C5 bond in the adenosine moiety as being too long. It generally
 refines to 1.46-1.47A. The quot;idealquot; distance in the
 validation report is 1.38A, and the upon review of the ATP.cif file in
 the REFMAC library, the target distance is 1.49A (and listed as a
 double bond). Clearly 1.37-1.38A is a reasonable target value. HIC-Up
 gives the target bond length as 1.404A.
   Where can I grab a revised ATP.cif file? I guess I'll need to re-refine
 all of my structures and re-run the validation report.BTW, I also
 looked at the PDB_REDO structure report for my structure, and can't
 reproduce the Rcryst and Rfree values with the same model.
   Bernie --

Re: [ccp4bb] 18A rule reference

[Santarsiero, Bernard D.]

Divide by 11 if you count all atoms, including H-atoms; divide by 20-22 if
you only count non-H-atoms, to get the approximately number of atoms in
the unit cell.


On Mon, May 5, 2014 6:07 am, Natalie Tatum wrote:
 Hi all,

 I've had an interesting question from an undergradute student, asking if
 there's any specific reference for the 18Å rule of thumb in small
 molecule crystallography for dividing the volume of a unit cell by 18 for
 the approximate number of non-hydrogen atoms.

 I'm afraid my google-fu has failed me, but I was curious as to whether
 there is a reference associated with the practice of if it's just one of
 those handy little pieces of advice for crystal screening.

 Thanks,

 Natalie

Re: [ccp4bb] Td ions in refmac

[Santarsiero, Bernard D.]

Is there a reason why you must use restraints with metal coordination?

The main reasons that you use restraints in protein refinement is to
reduce the number of free variables in the refinement, and because the
resolution isn't sufficient to resolve two atoms separated by the
typical carbon-carbon distance of 1.5A.

In a metal complex, the distances are roughly 2A or greater, and the metal
position is well-determined since it has a large number of electrons.  I
would just restrain the protein part of the structure, and let the metal
refine freely.

Bernie Santarsiero


On Fri, June 7, 2013 9:25 am, Roger Rowlett wrote:
 Well, when I solved a structure for a protein-rhenium adduct I had to
 manually code a restraint file for the rhenium tricarbonyl fragment,
 which is octahedral overall (with proteins ligands and waters), and had
 several different types of bond distances for the various ligands. I
 started with an incorrect description of a similar ligand cif file from
 HIC-UP and hand-edited the cif file to suit. It seems to me it should be
 possible (although somewhat painful) to write a similar restraint file
 (geometry + bond lengths) for ZN.

 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 6/7/2013 9:23 AM, Tim Gruene wrote:
 -BEGIN PGP SIGNED MESSAGE-
 Hash: SHA1

 Hi Jim,

 the chemical basis would be that the Zn connects to all four S atoms,
 and if they are indistinguishable, all Zn-S distances should be alike.

 But the S-atoms are not bonded to each other, hence they want to have as
 much distance from each other as possible. Since they are four, this
 makes them a tetrahedron.

 I am not aware of a more automated/ simpler way of turning the angles
 at the metal into restraints other than calculating the S-S distances
 with a pocket calculator, although that does not mean they do not exist.

 Cheers,
 Tim

 On 06/07/2013 11:56 AM, James Naismith wrote:
 Andrew Leslie made a similar suggestion for using CB to S
 distances.

 Both make me unhappy for the reason that there is not a clear
 chemical basis to them (Tim's has more a chemical basis) and how
 would I choose the distance. (I can measure it from a known
 structure)


 I have previously experimented with Tim's (works). I am searching
 for anyone who knew a simple way to restrain metal ion geometry in
 remfac based on angles at metal.

 best Jim








 On 7 Jun 2013, at 10:14, Tim Gruene t...@shelx.uni-ac.gwdg.de
 wrote:

 Dear James,

 I am not sure I fully picture what your coordination looks like,
 but my guess is that if you include distance restraints between the
 S-atoms you can distinguish between a tetrahedral geometry and a
 square pyramid.

 You could use external restraints in refmac - just make a dummy
 run with prosmat to get the syntax right and copy and modify a few
 lines tailored to your problem.

 Best, Tim

 On 06/07/2013 10:39 AM, James Naismith wrote:
 Thanks for this other suggestions similar (see  below.) My
 problem is not the distances per se, these restrain fine to
 2.3A

 I am relatively low resolution and my problem is with just
 distance restraints I move to a square planar arrangement. I
 suspect this is a feature of the geometry restraints
 dominating refinement (c'est la vie at low res). I do not
 force Zn to be anything (the cif file just has a distance) I
 think.

 What I need to do is create another restraint someone how
 that imposes a Td geometry.

 What I cannot figure how to do is set this up because it
 involves 5 residues (4 cys and 1 Zn).

 best Jim

 Jim Naismith BSRC, North Haugh j...@st-andrews.ac.uk

 Google scholar is free and tracks your outputs, mine are
 http://scholar.google.co.uk/citations?hl=enuser=fLmKKQMJ



 ResearcherID H-3408-2012; SCOPUS_ID 7005691850
 http://www.researchgate.net/profile/James_Naismith/

 The University of St Andrews is a charity registered in
 Scotland : No SC013532k




 On 6 Jun 2013, at 16:11, Roger Rowlett
 rrowl...@colgate.edu wrote:

 You can find instructions for restraining Zn-ligand
 distances here:

 http://capsicum.colgate.edu/chwiki/tiki-index.php?page=Model+Refinementhighlight=metal%20ion#Restraining_metal_ligand_distances_during_refinement_in_REFMAC






 Using this method, only Zn-ligand bond distances are restrained.
 REFMAC can usually find on its own certain Zn-atom linkages (I
 think it can find ZN-SG bonds with Cys by itself, but the version
 I am using usually misses or ZN-OD1 or ZN-OD2 bonds with Asp. This
 I pick up in the .cif file generated by REFMAC and include in a
 library file to be used for refinement. I have not had any issues
 with REFMAC altering ZN geometry. What you describe sounds like
 the geometry of the ZN atom has been defined or assumed somewhere
 in your approach.
 Cheers,

 

[ccp4bb] Phaser - changing number of clashes parameter

[Santarsiero, Bernard D.]

The new CCP4I interface for PHASER removed the menu item where you could
change the number of clashes. It's not set as a default to 5% of the
number of C-alpha atoms. Where is the default file?

I found the phaser_MR.def file in share/ccp4i/tasks, and changed the
parameter value, but it still used the default of 5%.

Thanks,

Bernie
-- 
Bernard D. Santarsiero
Research Professor
Center for Pharmaceutical Biotechnology and the
 Department of Medicinal Chemistry and Pharmacognosy
Center for Structural Biology
Center for Clinical and Translational Science
University of Illinois at Chicago
MC870  3070MBRB  900 South Ashland Avenue
Chicago, IL 60607-7173  USA
(312) 413-0339 (office)
(312) 413-9303 (FAX)
http://www.uic.edu/labs/bds

Re: [ccp4bb] Fun Question - Is multiple isomorphous replacement an obsolete technique?

[Santarsiero, Bernard D.]

Remember that it's all relative to the length of the FP vector. If your FP
vector is small, then the f component can substantially change the phase,
even with a small f component. So if you have measured a number of
relatively weak reflections with minimal error, there is a substantial
anomalous signal.  If you have a huge FP vector, then you won't see much
of a phase change.  Bernie


On Wed, June 6, 2012 10:02 am, Dyda wrote:
I suspect that pure MIR (without anomalous) was always a fiction. I doubt
 that anyone has ever used it. Heavy atoms always give
an anomalous signal

Phil

Re: [ccp4bb] Fun Question - Is multiple isomorphous replacement an obsolete technique?

[Santarsiero, Bernard D.]

No, I listed a few recent ones


V. Gaur, et al., Plant Physiol., 152(4), 1842-1850 (2010)

O. Antipova, J Biol Chem. 2010 Mar 5;285(10):7087-96. Epub 2010 Jan 6.

Y. Nakajima, J Bacteriol. 2008 Dec;190(23):7819-29. Epub 2008 Sep 26.

S. Stayrook, Nature. 2008 Apr 24;452(7190):1022-5.

Many MIRAS, so the MIR part helped to get forms, and then collected with AS.



On Wed, June 6, 2012 3:42 pm, Boaz Shaanan wrote:
 So if get the gist of the thread right, am I correct in assuming that the
 last protein structures to be solved strictly by MIR  are
 haemoglobin/myoglobin, lysozyme and chymotrypsin and perhaps one or two
 more in the late sixties? In which case the answer  to the original
 question about MIR being obsolete, is yes it is since a long time?

   Boaz

Re: [ccp4bb] Collecting small-molecule diffraction on a Macromolecular xtallography beam line

[Santarsiero, Bernard D.]

Most beamlines have attenuators, so there's little reason to collect
multiple sweeps. We always collect 360deg. Since it's a small molecule,
and usually fairly large and robust, you can warm it up, nudge it in a
different direction with a pin (we use sterile, disposable acupunture
needle), and refreeze it in the cryostream. Then do a second sweep in a
different orientation.

I recommend moving the beam energy to 15.5KeV or higher to compress the
diffraction image. Collect with 5-10deg widths. We can typically get the
detector to around 70-80mm. You need to get to 0.85A resolution or better
for good, stable refinement, and Acta Cryst. requires that resolution for
publication.  Often you need the low-resolution data and data to better
than 1A to help with the sigma2 relationships in direct methods.

You see both primary and secondary extinction, and that extinction can be
anisotropic, so the SWAT option in SHELX is most useful. Otherwise, the
overall scale factor is off, typically overestimated by the strong
low-resolution reflection intensities, with the result that the
anisotropic Gaussian displacement parameters may become non-positive
definate.

Bernie

On Wed, February 8, 2012 12:46 pm, Jens Kaiser wrote:
 Giorgo,
   We have done that routinely for quite some time now. We had problems
 when using a normal CCD detector, where we had to collect two or three
 sweeps to avoid overloads (see below). Since we have the PILATUS this is
 not necessary anymore and the data behaves fine. Problems still
 persisting are: we have only a single axis goniometer, which can lead to
 low completeness in P1 and P-1. Highest energy (17keV) and closest
 distance (188mm) at our beamline have many SM crystals (even the ones
 that don't diffract in house -- that is a 300 or 500 u sealed tube)
 with an I/sig of 5-10 at the edge of the detector. Crunch, Acorn,
 ShelxCDE and ShelxS don't have any problem with any of the data we
 collected to 0.9A resolution. The multipass caused some inexplicable
 non definite positives during refinement. We haven't tracked that down
 systematically, so it might just have happened haphazardly.

 HTH,

 Jens

 On Wed, 2012-02-08 at 11:41 +, Giorgio Giardina wrote:
 Hello,
 I have some interesting small molecule xtals.
 I was wondering if it is possible to collect a small molecule data-set
 using a sincrotron macromolecular  xtallography beam line, maybe with a
 very low beam intensity and moving the detector as close as possible?
 Has anybody experienced that?
 And if I get the images back home,  can I process them using standard
 macromolecular software or do I need ab-initio special programs?
 Will MR work for phasing?

 Thanks in advance,
 Giorgio




-- 
Bernard D. Santarsiero
Research Professor
Center for Pharmaceutical Biotechnology and the
 Department of Medicinal Chemistry and Pharmacognosy
Center for Structural Biology
Center for Clinical and Translational Science
University of Illinois at Chicago
MC870  3070MBRB  900 South Ashland Avenue
Chicago, IL 60607-7173  USA
(312) 413-0339 (office)
(312) 413-9303 (FAX)
http://www.uic.edu/labs/bds

[ccp4bb] Coot, RH6.1 x86 linux

[Santarsiero, Bernard D.]

I just installed CCP4-6.2.0 and COOT on a Red Hat v6.1, x86, linux
workstation. Some of the menus, along the right edge, are gray buttons on
gray background. Where is the preferences file to change the colors of the
menu buttons? The menus across the top of the window are fine.

Also, idiffdisp is unable to view a file. It loads the idiffdisp window,
but gives error messages. Is there a log file that I can scan? I think I'm
missing some library files for x86.

Bernie

-- 
Bernard D. Santarsiero
Research Professor
Center for Pharmaceutical Biotechnology and the
 Department of Medicinal Chemistry and Pharmacognosy
Center for Structural Biology
Center for Clinical and Translational Science
University of Illinois at Chicago
MC870  3070MBRB  900 South Ashland Avenue
Chicago, IL 60607-7173  USA
(312) 413-0339 (office)
(312) 413-9303 (FAX)
http://www.uic.edu/labs/bds

Re: [ccp4bb] Newbie Installation Question

[Santarsiero, Bernard D.]

I've had problems too. Some of the files are non-readable, so you need to
do a

sudo chmod -R a+r *

to make them readable. Also it looks like the libraries changed a bit (at
least for ADP).

Bernie

On Fri, July 29, 2011 2:08 pm, Yuri wrote:
 Dear all,
 I have just downloaded and installed the ccp4-6.2.0.
 It says all I should do next is source the /setup-scripts/csh/ccp4.setup
 file... I have done that, but I cannot launch the program...
 Any help is welcome...(it is probably something really stupid on my
 part...)
 Best,

Re: [ccp4bb] small lysozyme crystals?

[Santarsiero, Bernard D.]

That's a really old paper. You can purchase the lysozyme from Hampton
Research and it's fine. The recipe is available from the Hampton Research
page:

http://hamptonresearch.com/product_detail.aspx?cid=28sid=173pid=524

Grow them a low temp and you can stop them when they are the right size. 
I favor that over room temp. They grow fast and large, but don't give a
good R(merge) as when grown at a lower concentration and slower.

Bernie



On Tue, July 26, 2011 5:09 pm, mjvdwo...@netscape.net wrote:


  James,

 I would have a look at the paper by Judge et al:

 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1300446/pdf/10465769.pdf


 Specifically, in this paper you will find that the crystallization
 behavior of lysozyme changes drastically with pH. At the time the paper
 wasn't really written to manipulate for small crystal size, but looking
 back at the paper (specifically Fig 5), it appears that you can read the
 conditions that will give you crystals around the size you want.

 Not re-reading the paper, quoting from memory (which we all think is
 better than it really is), it is important to use good quality lysozyme to
 get reproducible results. Good quality probably means freshly purified
 from fresh (farm-acquired) eggs. I am not kidding you, it makes a big
 difference. Also, I am going out on a limb to say (I know you know this)
 that the buffer preparation method matters a lot. Taking sodium acetate
 solution and pH-ing it with HCl will give very different results from
 taking acetic acid and pH-ing it with NaOH (because the ionic strength of
 the buffer is not the same). Lysozyme crystallizes so easily that we tend
 to forget tedious details.

 Hope this helps. This paper will probably give you some ideas in the right
 direction.

 Mark van der Woerd

Re: [ccp4bb] problem of conventions

[Santarsiero, Bernard D.]

Dear Ian,

Well, it *IS* broke. If you are running some type of process, as you
implied in referring to LIMS, then there is a step in which you move from
the crystal system and point group to the actual space group. So, at that
point you identify P22121. The next clear step, automatically by software,
is to convert to P21212, and move on. That doesn't take an enormous amount
of code writing, and you have a clear trail on how you got there.

To be even more intrusive, what if you had cell parameters of 51.100,
51.101, and 51.102, and it's orthorhombic, P21212. For other co-crystals,
soaks, mutants, etc., you might have both experimental errors and real
differences in the unit cell, so you're telling me that you would process
according to the a  b  c rule in P222 to average and scale, and then it
might turn out to be P22121, P21221, or P21212 later on? When you wish to
compare coordinates, then you have re-assign one coordinate data to match
the other by using superposition, rather than taking on an earlier step of
just using the conventional space group of P21212?

Again, while I see use of the a  b  c rule when there isn't an
overriding reason to assign it otherwise, as in P222 or P212121, there
*is* a reason to stick to the convention of one standard setting. That's
the rationale on using P21/n sometimes vs. P21/c, or I2 vs C2, to avoid a
large beta angle, and adopt a non-standard setting.

Finally, if you think it's fine to use P22121, then can I assume that you
also allow the use of space group A2 and B2?

Bernie


Bernie







On Fri, April 1, 2011 8:46 am, Ian Tickle wrote:
 Dear Gerard,

 The theory's fine as long as the space group can be unambiguously
 determined from the diffraction pattern.  However practice is
 frequently just like the ugly fact that destroys the beautiful theory,
 which means that a decision on the choice of unit cell may have to be
 made on the basis of incomplete or imperfect information (i.e.
 mis-identification of the systematic absences).  The 'conservative'
 choice (particularly if it's not necessary to make a choice at that
 time!) is to choose the space group without screw axes (i.e. P222 for
 orthorhombic).  Then if it turns out later that you were wrong it's
 easy to throw away the systematic absences and change the space group
 symbol.  If you make any other choice and it turns out you were wrong
 you might find it hard sometime later to recover the reflections you
 threw away!  This of course implies that the unit-cell choice
 automatically conforms to the IT convention; this convention is of
 course completely arbitrary but you have to make a choice and that one
 is as good as any.

 So at that point lets say this is the 1970s and you know it might be
 several years before your graduate student is able to collect the
 high-res data and do the model-building and refinement, so you publish
 the unit cell and tentative space group, and everyone starts making
 use of your data.  Some years later the structure solution and
 refinement is completed and the space group can now be assigned
 unambiguously.  The question is do you then revise your previous
 choice of unit cell risking the possibility of confusing everyone
 including yourself, just in order that the space-group setting
 complies with a completely arbitrary 'standard' (and the unit cell
 non-conventional), and requiring a re-index of your data (and
 permutation of the co-ordinate datasets).  Or do you stick with the IT
 unit cell convention and leave it as it is?  For me the choice is easy
 ('if it ain't broke then don't fix it!').

 Cheers

 -- Ian

 On Fri, Apr 1, 2011 at 1:40 PM, Gerard Bricogne g...@globalphasing.com
 wrote:
 Dear Boaz,

     I think you are the one who is finally asking the essential
 question.

     The classification we all know about, which goes back to the 19th
 century, is not into 230 space groups, but 230 space-group *types*, i.e.
 classes where every form of equivalencing (esp. by choice of setting)
 has
 been applied to the enumeration of the classes and the choice of a
 unique
 representative for each of them. This process of maximum reduction
 leaves
 very little room for the introducing conventions like a certain
 ordering
 of the lengths of cell parameters. This seems to me to be a major
 mess-up in
 the field - a sort of second-hand mathematics by (IUCr) committee
 which
 has remained so ill-understood as to generate all these confusions. The
 work
 on the derivation of the classes of 4-dimensional space groups explained
 the
 steps of this classification beautifully (arithmetic classes -
 extension by
 non-primitive translations - equivalencing under the action of the
 normaliser), the last step being the choice of a privileged setting *in
 termns of the group itself* in choosing the representative of each
 class.
 The extra convention abc leads to choosing that representative in a
 way
 that depends on the metric properties of the sample instead of once and
 for
 all 

Re: [ccp4bb] problem of conventions

[Santarsiero, Bernard D.]

If you are using CCP4, it can accomodate P22121. However, just reindex in
CCP4 to the correct setting with P21212.

Bernie Santarsiero


On Thu, March 31, 2011 9:28 am, Anita Lewit-Bentley wrote:
 Dear all,

 I would like to share my experiencde with a rather unexpected problem
 of indexing conventions. Perhaps I can save people some time

 I have a crystal in the more unusual P21212 space-group (No 18). Its
 unit cell lengths are bac (please note). I systematically use XDS
 for data integration, since so far it was able to handle even the most
 horrible-looking spots.

 Now XDS indexed my data in space-group 18, but with the axes order
 abc! It had, in fact, invented a space-group P22121, which does
 not exist. I did not realise this until I had spent a couple of weeks
 with beautiful peaks in rotation functions, but hopeless results in
 translation functions. It wasn't until I looked more closely into the
 definition of the screw axes that I realised the problem.

 POINTLESS does not allow a reindexing of reflexions within the same
 space-group, but fortunately REINDEX did the trick at the level of
 intensities, because I like to use SCALA for careful scaling of my data.

 So, basically, beyond just warning people who might encounter similar
 problems, I was wo,dering if XDS could perhaps reindex reflexions
 according to Int. Table conventions once the screw axes of a crystal
 system have been identified?

 With best wishes,

 Anita


 Anita Lewit-Bentley
 Unité d'Immunologie Structurale
 CNRS URA 2185
 Département de Biologie Structurale  Chimie
 Institut Pasteur
 25 rue du Dr. Roux
 75724 Paris cedex 15
 FRANCE

 Tel: 33- (0)1 45 68 88 95
 FAX: 33-(0)1 40 61 30 74
 email: ale...@pasteur.fr

Re: [ccp4bb] problem of conventions

[Santarsiero, Bernard D.]

Interesting. My IT, both volume I and volume A (1983) only have P21212 for
space group #18. Do I have to purchase a new volume A every year to keep
up with the new conventions?

Cheers,

Bernie


On Thu, March 31, 2011 12:57 pm, Ian Tickle wrote:
 I would like to share my experiencde with a rather unexpected problem of
 indexing conventions. Perhaps I can save people some
 time

 I have a crystal in the more unusual P21212 space-group (No 18). Its
 unit cell lengths are bac (please note). I systematically
 use XDS for data integration, since so far it was able to handle even
 the most horrible-looking spots.

 Now XDS indexed my data in space-group 18, but with the axes order
 abc! It had, in fact, invented a space-group P22121,
 which does not exist. I did not realise this until I had spent a couple
 of weeks with beautiful peaks in rotation functions, but
 hopeless results in translation functions. It wasn't until I looked more
 closely into the definition of the screw axes that I realised the
 problem.

 POINTLESS does not allow a reindexing of reflexions within the same
 space-group, but fortunately REINDEX did the trick at the
 level of intensities, because I like to use SCALA for careful scaling of
 my data.

I was wo,dering if XDS could perhaps reindex reflexions according
 to Int. Table conventions once the screw axes of a crystal system have
 been
 identified?

 The International Tables / IUCr / NIST convention _is_  a=b=c for
 orthorhombic so no re-indexing is necessary or desirable.  See IT vol.
 A 5th ed. (2002), table 9.3.4.1 (p. 758 in my edition) for all the
 conventional cells.  The problem may be that some programs are not
 sticking to the agreed convention - but then the obvious solution is
 to fix the program (or use a different one).  Is the problem that XDS
 is indexing it correctly as P22121 but calling it SG #18 (i.e. instead
 of the correct #3018).  That would certainly confuse all CCP4 programs
 which generally tend to use the space-group number first if it's
 available.

 I'm not clear what you mean when you say P22121 doesn't exist?  It's
 clearly shown in my edition of IT (p. 202).  Maybe your lab needs to
 invest in the most recent edition of IT?

 Cheers

 -- Ian

Re: [ccp4bb] problem of conventions

[Santarsiero, Bernard D.]

Ian,

I think it's amazing that we can program computers to resolve a  b  c
but it would be a major undertaking to store the matrix transformations
for 22121 to 21212 and reindex a cell to a standard setting. I was also
told that I was lazy to not reindex to the standard setting when I was a
grad student. Now it takes less than a minute to enter a transformation
and re-index.

The orthorhombic rule of a  b  c makes sense in 222 or 212121, but when
there is a standard setting of the 2-fold along the c-axis, then why not
adopt that? Often we chose a non-setting when there was a historical
precedence, as in the comparison of one structure to another, e.g., P21/c
with beta greater than 120deg vs. P21/n, etc. That is no more difficult
with modern computing than dragging along three space groups for #18.
There was a compactness to 230, and only 230 space groups. (I cheat, since
I agree there is both the rhombohedral and hexagonal cell settings for
R3bar.)

Bernie



On Thu, March 31, 2011 5:48 pm, Ian Tickle wrote:
 On Thu, Mar 31, 2011 at 10:43 PM, James Holton jmhol...@lbl.gov wrote:
 I have the 2002 edition, and indeed it only contains space group
 numbers up to 230.  The page numbers quoted by Ian contain space group
 numbers 17 and 18.

 You need to distinguish the 'IT space group number' which indeed goes
 up to 230 (i.e. the number of unique settings), from the 'CCP4 space
 group number' which, peculiar to CCP4 (which is why I called it
 'CCP4-ese'), adds a multiple of 1000 to get a unique number for the
 alternate settings as used in the API.  The page I mentioned show the
 diagrams for IT SG #18 P22121 (CCP4 #3018), P21221 (CCP4 #2018) and
 P21212 (CCP4 #18), so they certainly are all there!

 Although I am all for program authors building in support for the
 screwy orthorhombics (as I call them), I should admit that my
 fuddy-duddy strategy for dealing with them remains simply to use space
 groups 17 and 18, and permute the cell edges around with REINDEX to
 put the unique (screw or non-screw) axis on the c position.

 Re-indexing is not an option for us (indeed if there were no
 alternative, it would be a major undertaking), because the integrity
 of our LIMS database requires that all protein-ligand structures from
 the same target  crystal form are indexed with the same (or nearly
 the same) cell and space group (and it makes life so much easier!).
 With space-groups such as P22121 it can happen (indeed it has
 happened) that it was not possible to define the space group correctly
 at the processing stage due to ambiguous absences; indeed it was only
 after using the SGALternative ALL option in Phaser and refining each
 TF solution that we identified the space group correctly as P22121.

 Having learnt the lesson the hard way, we routinely use P222 for all
 processing of orthorhombics, which of course always gives the
 conventional abc setting, and only assign the space group well down
 the pipeline and only when we are 100% confident; by that time it's
 too late to re-index (indeed why on earth would we want to give
 ourselves all that trouble?).  This is therefore totally analogous to
 the scenario of yesteryear that I described where it was common to see
 a 'unit cell' communication followed some years later by the structure
 paper (though we have compressed the gap somewhat!), and we base the
 setting on the unit cell convention for exactly the same reason.

 It's only if you're doing 1 structure at a time that you can afford
 the luxury of re-indexing - and also the pain: many times I've seen
 even experienced people getting their files mixed up and trying to
 refine with differently indexed MTZ  PDB files (why is my R factor so
 high?)!  My advice would be - _never_ re-index!

 -- Ian


  I have
 yet to encounter a program that gets broken when presented with data
 that doesn't have abc, but there are many non-CCP4 programs out
 there that still don't seem to understand P22121, P21221, P2122 and
 P2212.

 I find that surprising!  Exactly which 'many' programs are those?  You
 really should report them to CCP4 (or to me if it's one of mine) so
 they can be fixed!  We've been using CCP4 programs as integral
 components of our processing pipeline (from data processing through to
 validation) for the last 10 years and I've never come across one
 that's broken in the way you describe (I've found many broken for
 other reasons and either fixed it myself or reported it - you should
 do the same!).  Any program which uses csymlib with syminfo.lib can
 automatically handle all space groups defined in syminfo, which
 includes all the common alternates you mentioned (and others such as
 I2).  The only program I'm aware of that's limited to the standard
 settings is sftools (because it has its own internal space group table
 - it would be nice to see it updated to use syminfo!).

 This is not the only space group convention issue out there!  The
 R3x vs H3x business continues to be annoying to this day!

Re: [ccp4bb] X-Ray films

[Santarsiero, Bernard D.]

This was mentioned in Marv Hackert's outgoing remarks at the ACA meeting a
few years ago, and he showed the clip of the scene. Jimmy Stewart's
comment is apt: It's quite fundamental. It's really odd to me that the
schools don't teach it.

Bernie


On Sun, March 27, 2011 4:13 pm, harry powell wrote:
 Yes, this is based on a Neville Shute story, stars James Stewart and
 Marlene Dietrich.

 James Stewart's character's daughter is interested in
 crystallography, and her name is Elspeth (a BCA connection?)

 On 27 Mar 2011, at 21:04, Frances C. Bernstein wrote:

 Picking up an old thread...

 Herbert and I just saw the 1951 film No Highway in the Sky
 and it has several mentions of crystallography.

 Frances

 =
 Bernstein + Sons
 *   *   Information Systems Consultants
 5 Brewster Lane, Bellport, NY 11713-2803
 *   * ***
  *Frances C. Bernstein
   *   ***  f...@bernstein-plus-sons.com
  *** *
   *   *** 1-631-286-1339FAX: 1-631-286-1999
 =

 Harry
 --
 Dr Harry Powell,
 MRC Laboratory of Molecular Biology,
 Hills Road,
 Cambridge,
 CB2 0QH

Re: [ccp4bb] CNS and protein structure refinement

[Santarsiero, Bernard D.]

Ditto.  YES, still generally prefer CNS v1.3 to REFMAC, especially due to
composite and simulated-annealing maps.

Different set of flagged reflections, so it might very well be different.
Also, the low-resolution solvent scaling is different between the two
refinements.

Bernie Santarsiero



On Wed, February 16, 2011 7:53 am, Boaz Shaanan wrote:
 Yes X 2

 Boaz

 - Original Message -
 From: REX PALMER rex.pal...@btinternet.com
 Date: Wednesday, February 16, 2011 15:32
 Subject: [ccp4bb] CNS and protein structure refinement
 To: CCP4BB@JISCMAIL.AC.UK

 Does anyone still use CNS ?
 Do we expect Rfree from CNS for example to be different from the
 value given by Refmac at the end of the refinement?
  
 Rex Palmer
  
 Birkbeck College

 Boaz Shaanan, Ph.D.
 Dept. of Life Sciences
 Ben-Gurion University of the Negev
 Beer-Sheva 84105
 Israel
 Phone: 972-8-647-2220 ; Fax: 646-1710
 Skype: boaz.shaanan‎

Re: [ccp4bb] From oscillation photographs to seeing specific sections of reciprocal lattice.

[Santarsiero, Bernard D.]

I think you should just grab a copy of Stout and Jensen, and use the
oscillation photographs directly. What's so special about a precession
image? You can still index the spots and follow along rl lines.

Bernie Santarsiero



On Thu, June 25, 2009 1:38 pm, Francis E Reyes wrote:
 Yes this is exactly what I wanted. I'm embarking on an educational
 pursuit of determining the space group from the diffraction images
 directly. Unfortunately  it seems like all the solutions insofar are
 only commercially available as part of large packages that don't list
 their prices directly on the website and, therefore, are probably too
 much for a single person to afford for just this purpose.

 Cheers

 FR



 On Jun 25, 2009, at 10:41 AM, Ian Tickle wrote:

 But I thought what you wanted was to reconstruct the diffraction
 pattern
 (i.e. streaks, TDS, ice rings, zingers, warts  all) as a
 pseudo-precession image, not just display a representation of the
 integrated intensities.  That surely would be much more useful, then
 one
 could see whether the apparent systematic absence violations were just
 streaks from adjacent spots, TDS, ice spots etc that have fooled the
 integration algorithm.  That would be much more useful!  In the days
 when we had real precession cameras this was how you assigned the
 space
 group.

 Cheers

 -- Ian

 -
 Francis Reyes M.Sc.
 215 UCB
 University of Colorado at Boulder

 gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D

 8AE2 F2F4 90F7 9640 28BC  686F 78FD 6669 67BA 8D5D

Re: [ccp4bb] peak height at mouse click

[Santarsiero, Bernard D.]

Empirically, you can leave out a couple of average atoms in the
structure, and recalculate the maps. If you leave out a O, N, and C atom,
with relatively average B's, then you know how many electrons you should
be seeing for each in the difference map. Note that the peak height will
vary due to errors in phasing or B-factor.

Bernie Santarsiero



On Mon, April 20, 2009 11:18 am, Pavel Afonine wrote:
 Hi Thiyagarajan,

 the latest version of PHENIX has a command line tool called
 phenix.real_space_correlation, that for each atom or residue computes:

 - map CC (where a user can define any map types, the default: 2mFo-DFc
 and Fc maps),
 - 2mFo-DFc map value at atom center,
 - mFo-DFc map value at atom center.

 Please let me know if you have question,
 Pavel.


 On 4/20/09 3:47 AM, S. Thiyagarajan wrote:
 Dear CCP4 users

 Is there any easy way of calculating the peak height / number of
 electrons at a given position, say a mouse click point in coot.

 Is there any formula to calculate the number of electrons based on
 sigma level and peak height, as given in difference map peaks in coot.

 I have some peaks in my map which take water or sodium/magnesium or
 chlorine atom with out giving out any positive or negative density
 upon further refinement.

 The asymmetric unit has about 425 residues and the data resolution is
 1.5A.

 Thanks and regards

 S. Thiyagarajan
 Department of Cell and Organism Biology
 Lund University
 Sölvegatan 35
 Lund, Sweden

Re: [ccp4bb] Halide soaking

[Santarsiero, Bernard D.]

We were on a Br-soak flurry for awhile, and when I also talked to others,
it would work about 20-30% of the time. A few years I compared a structure
with Se-Met and native with Br-soak. There were four Br's at about 30%
occupancy, and it phased fine. It wasn't as good as the Se-Met phasing,
but we could trace and build the structure independently. It's so easy to
try that my suggestion has been to try it out first, and if it works, then
great. Also, if I had redone it, I would have soaked for longer than 20-30
seconds to try and increase the occupancy.

Bernie Santarsiero


On Tue, March 31, 2009 12:17 pm, Ethan Merritt wrote:
 On Tuesday 31 March 2009 09:57:13 Jose Antonio Cuesta-Seijo wrote:
 Hi!

 Normally the cell parameters, etc change very very little. You'll
 only know if the bromides got in at the synchrotron by looking at the
 fluorescence spectrum

 That won't help, normally.  It only tells you that there is bromide in
 the solution, not that it found ordered positions in your crystal.

 and at the anomalous signal. Normally some will
 make it in and some will be in ordered sites, then it becomes mostly
 a question of data quality to detect it.

 Right.  You have to process the data and look for a real signal.
 The mere presence of Br is not enough.

 In my experience, the frustrating thing is that even if the Br soak
 works in the sense that it introduces Br into your lattice, the
 signal is often/usually not sufficient to phase the structure
 de novo.  Nevertheless, when you eventually do solve and refine
 the structure, you can go back to your anomalous difference data
 and calculate a Bijvoet Difference Fourier that clearly shows the
 Br sites.

   Ethan


 You could also try the equivalent iodide soak. Iodine has a decent
 anomalous signal at the copper wavenght and thus the anomalous signal
 can be detected at your home source and many times the structure can
 be solved by SAD or SIRAS. I would also thing that conditions that
 give ordered iodide sites are likely to result in ordered bromide
 sites, although the ions are not identical.

 Jose.


 **
 Jose Antonio Cuesta-Seijo
 Cancer Genomics and Proteomics
 Ontario Cancer Institute, UHN
 MaRS TMDT Room 4-902M
 101 College Street
 M5G 1L7 Toronto, ON, Canada
 Phone:  (416)581-7544
 Fax: (416)581-7562
 email: jcue...@uhnres.utoronto.ca
 **



 On Mar 31, 2009, at 12:19 PM, tat cheung cheng wrote:

  Hi all
 
  I am now trying to do bromide soaking, but i am not really sure
  does the bromide atom enter my crystal. So is there any signs that
  indicate the entry of bromide atom? e.g. does the space group, cell
  dimension change? or just nothing change, and the bromide atom just
  get in?
  Thanks very much.
 
  T.C. Cheng
 
 
Yahoo!香港提供網上安全攻略，教你如何防範黑客!
 請前往 http://
  hk.promo.yahoo.com/security/ 了解更多!
 







 --
 Ethan A Merritt
 Biomolecular Structure Center
 University of Washington, Seattle 98195-7742

Re: [ccp4bb] X-Stream 2000 problem - ICING

[Santarsiero, Bernard D.]

We've been able to run months with an old Xstream 2000 system, so that
shouldn't be the problem. Unlike Frank, we haven't had problems with water
in the nitrogen from a nitrogen generator.

If Frank is correct, that it's water, then either the molecular sieves
need to be replaced, or there is ice buildup and blockage in the coldhead.
Usually blockage in the coldhead means that you can't get down in temp, or
can't maintain it within 1 degree or less. We do bring the temp up, run it
at RT over the weekend to dry things out, and then bring it down for
routine cooling.

If it's turbulence, then you don't have the warm and cold flow rates set
correctly. One thing you can see immediately, upon getting down to 100K,
is how far is the frosting plume from the tip of the nozzle. If you adjust
the flow rates, you can move that plume, caused by turbulence, closer or
further from the nozzle. It's also possible that there is something in the
nozzle that is causing turbulence no matter what the flow rates are, and
your vendor needs to look at that.

Bernie Santarsiero


On Mon, January 12, 2009 6:11 am, Frank von Delft wrote:
 Hi Mark

 We had a LOT of pain with icing, and it really comes down to one thing:
 water in the gN2.  And don't expect to measure some other way whether
 you have it, because your X-stream (or Cobra) is the most sensitive
 water meter there is.

 In our case, the symptom was the X-stream (and later Cobra) blocking up
 after between several days and several hours.  And we solved it by
 ditching the gN2 generators we were using, and organising boil-off gN2
 (much purer).

 The secondary effect of sample icing:  we'd see this as well if we had
 something (e.g. collimator) poking into the cold stream*:  that causes
 turbulence which draws in moisture.  Worst case you see ice flakes flick
 onto the crystal in real time;  best case you get an ice ball after a
 few minutes to hours.

 (* technically, the stationary phase between the cold and warm streams.)

 Hope that helps.
 phx


 Mark Agacan wrote:
 Apologies for this slightly off topic question:

 I am having a great deal of trouble with my X-Stream 2000 cryostream
 system and I wondered if other users have similar problems.

 I've replaced almost all components (new GAST compressors, helium
 recharges, filters, etc., etc.) in the last couple of months but there
 is almost always icing of any cryo within 10 - 20 minutes of mounting a
 loop, and it is adversely affecting data collections.

 It appears like there is too much moisture in the cold or wam streams
 but the tubes have been fully dried out as per Rigaku advice.

 This X-Stream is attached to a generator with inverted phi axis and and
 i'm wondering if this could be the source of the problem, as the
 X-Stream for another generator in the same laboratory with normal phi
 axis does not ice up.  Can some sort of turbulence around the loop
 caused by backdraft from the cryo hitting the inverted phi axis / camera
 mount cause excess humidity and lead to icing on the pin, loop and
 crystal?

 Has anyone else got this problem?  Any suggestions would be very
 gratefully appreciated.

 Best Wishes,

 Mark


 _
 Dr Mark Agacan
 Scientific Officer,
 Division of Biological Chemistry
 and Drug Discovery,
 Wellcome Trust Biocentre,
 College of Life Sciences,
 Dow St.,
 University of Dundee,
 Dundee, DD1 5EH
 Tel: +44 1382 388751
 Fax: +44 1382 345764
 _
 The University of Dundee is a registered Scottish charity, No: SC015096

[ccp4bb] O/T: can a protein which dimerizes in solution crystallize as a monomer?

[Santarsiero, Bernard D.]

In parallel with the discussion around this off-CCP4-topic, are they any
good examples of the opposite case, where the protein is a monomer in
solution (as evident from light scattering, MW determination through
centrifugation, EPR, etc.) but crystallizes as a dimer or higher multimer?

Bernie Santarsiero

Re: [ccp4bb] Progresss with Stereo 3D under Mac OS X Leopard

[Santarsiero, Bernard D.]

My use goes back to Evans and Sutherland workstations, FRODO, SGI's with
O, Duncan's graphics program suite, and COOT, on multiple platforms.

I don't use dials at all, and multiple uses with a three-button mouse are
far more favorable.

I still prefer O to COOT for a number of reasons. The main reason is that
the graphics isn't as good on most PC/linux/mac boxes, the screen isn't as
large, and the color range isn't as good.

I rarely, if ever, use stereo options, even when available. You can in 3D
fine without it, and you need a system that has a fast enough refresh rate
and no shadowing. When users learn by building without stereo, they don't
see the advantages of it later on.

Bernie Santarsiero

 =
 Greetings.  This is a semi-informal survey of recent crystallography
 workstation users.  Please take a minute to respond.  Your answers will
 help us improve the crystallography computing environment.


 1) Have you recently (past few months) used a crystallography
 workstation
   for molecular model building and/or visualization?  YES  NO

   Answer:  YES


 2) If yes to (1), which model building software did you use (list all
   that apply)?  COOT  O  OTHER - please specify

   Answer:  O


 3) When model building, do you use the dial box?
   ALWAYS  OFTEN  SOMETIMES  RARELY  NEVER

   Answer:  NEVER


 4) When model building, do you use 3D stereo visualization (i.e. stereo
   glasses)?  ALWAYS  OFTEN  SOMETIMES  RARELY  NEVER

   Answer: RARELY to NEVER


 5) If yes to (1), which molecular visualization software did you use
 (list
   all that apply)?  COOT  O  CHIMERA  PYMOL  OTHER - please specify

   Answer:  PYMOL, O, RASMOL, INSIGHT


 6) When visualizing molecular models, do you use the dial box?
   ALWAYS  OFTEN  SOMETIMES  RARELY  NEVER

   Answer: NEVER


 7) When visualizing molecular models, do you use 3D stereo
 visualization
   (i.e. stereo glasses)?  ALWAYS  OFTEN  SOMETIMES  RARELY  NEVER

   Answer: NEVER


 8) Is there any software you would like to have available in the
   computing environment to assist you in molecular model building
 and/or
   visualization that is not currently available?

   Answer:  NO


 Thank you for your time.




 --
 Dr. Jeroen R. Mesters
 Gruppenleiter Strukturelle Neurobiologie und Kristallogenese
 Institut für Biochemie, Universität zu Lübeck
 Zentrum für Medizinische Struktur- und Zellbiologie
 Ratzeburger Allee 160, D-23538 Lübeck
 Tel: +49-451-5004070, Fax: +49-451-5004068
 Http://www.biochem.uni-luebeck.de
 Http://www.iobcr.org
 Http://www.selfish-brain.org
 Http://www.opticryst.org
 --
 If you can look into the seeds of time and say
 which grain will grow and which will not - speak then to me  (Macbeth)
 --




 --
 Scott D. Pegan, Ph.D.
 Senior Research Specialist
 Center for Pharmaceutical
 Biotechnology
 University of Illinois at Chicago

Re: [ccp4bb] Friedel vs Bijvoet

[Santarsiero, Bernard D.]

Friedel pair is strictly F(hkl) and F(-h,-k,-l).
Bijvoet pair is F(h) and any mate that is symmetry-related to F(-h), e.g.,
F(hkl) and F(-h,k,-l) in monoclinic.

There are always anomalous differences, though they can be unmeasurably
small.

Bernie Santarsiero

On Thu, June 26, 2008 10:55 am, Bernhard Rupp wrote:
 Dear All,

 I wonder about the conventions using Friedel vs Bijvoet pair.

 a) there are no differences. As long as h = -h, it's a Friedel
or a Bijvoet pair. They are the same.

 b) A Friedel pair is any reflection h = -h including hR = -h, i.e.
including centric reflections.
A Bijvoet pair is an acentric Friedel pair, it can carry
anomalous amplitude differences, whereas centric Friedel
pairs invariably cannot. Actually, Bijvoet pairs (acentric
Friedel pairs) invariably do carry anomalous amplitude differences.
There is no such thing as no anomalous scattering.
We may elect to ignore it, only.

 c) of course, this all assumes absence of anisotropic AS.

 def b) seems to be helpful in discussions and make sense given that
 absolute

 configuration that needs AS signal is somehow associated with
 Bijvoet's
 work.

 Are any authoritative answers/conventions/opinions available on that ?

 Thx, BR

 -
 Bernhard Rupp
 001 (925) 209-7429
 +43 (676) 571-0536
 [EMAIL PROTECTED]
 [EMAIL PROTECTED]
 http://www.ruppweb.org/
 -
 The hard part about playing chicken
 is to know when to flinch
 -

Re: [ccp4bb] 2mhb

[Santarsiero, Bernard D.]

Bernhard,

It's complicated. The last two columns are related to atomic Z number and
group occupancy. They used a different equation than the typical
Debye-Waller equation to calculate a sort of B-factor and get the electron
density maps.

Bernie Santarsiero


On Fri, June 20, 2008 4:05 pm, Bernhard Rupp wrote:
 Dear All -

 can someone please explain to me the last 2 dat items in the
 2mhb ATOM recs? Looks like X Y Z followed by B and ??
 Is this some historic legacy (1977) or another remediation incident?
 Obviousy this is not the standard format it purports to be in,
 but there might be good reasons. No access to orig citation...

 Thx, br
 -
 Bernhard Rupp
 001 (925) 209-7429
 +43 (676) 571-0536
 [EMAIL PROTECTED]
 [EMAIL PROTECTED]
 http://www.ruppweb.org/
 -
 The hard part about playing chicken
 is to know when to flinch
 -

Re: [ccp4bb] is it Ok to freeze

[Santarsiero, Bernard D.]

I typically collect data at -50C on all small molecule samples. I've had
quite a few cases where there are phase transitions, and you can damage
the crystals, especially when the molecules are packed in a pi-pi stacking
motif, or I'm dealing with alloy systems.

I've also collected data at 16K, so it all depends on your sample.

Instead of finding out if there is a phase transition, -50C seems to be a
good choice of a temperature to reduce the displacement amplitudes,
radiation damage, and solvent loss.

Bernie Santarsiero


On Thu, June 19, 2008 9:40 am, Diana Tomchick wrote:
 Every small molecule dataset I collected as a graduate student in
 chemistry back in the mid to late 1980's was at 100K. I never had to
 worry about crystal slippage during collection, organic solvent
 evaporation, air oxidation of the sample (organometallic metal
 clusters) or secondary radiation damage.

 When I switched to protein crystallography, I was absolutely amazed
 when told that you can not cool a protein crystal below 4 degrees C
 for data collection.

 How times have changed,

 Diana

 On Jun 19, 2008, at 9:03 AM, Ian Tickle wrote:

 I would go along with Harry  friends, I used crystal cooling when I
 was
 at Aafje Vos' Struktuurchemie lab in Groningen in 1972, when the
 technique had already been in routine use there for at least 10 years,
 in order to study compounds that are liquid at ambient temp (of course
 it was custom-built kit using a collection of liq N2 Dewar vessel 
 tubes, nothing as fancy as a Cryostream!).  The Groningen group really
 pioneered the use of low temp for small molecule structures and I
 don't
 recall increased mosaicity ever being an issue.  Occasionally you
 would
 get a compound with a phase transition on the way down and the crystal
 would literally explode in a puff of powder before your eyes!  The
 motive for using low temp was of course to reduce the thermal motion
 and
 libration effects, and thus greatly improve the accuracy of the
 molecular geometry, and low temp is pretty well essential if you're
 into
 valence density deformation maps, again in order the minimise the
 contribution from thermal motion.

 -- Ian

 -Original Message-
 From: [EMAIL PROTECTED]
 [mailto:[EMAIL PROTECTED] On Behalf Of harry powell
 Sent: 19 June 2008 14:05
 To: Remy Loris
 Cc: CCP4BB@JISCMAIL.AC.UK
 Subject: Re: [ccp4bb] is it Ok to freeze

 Hi

 Without wishing to start an argument, I've been checking with
 some of my colleagues who are chemical crystallographers -
 the reply I get is that, for routine structural analysis,
 pretty well all datasets are collected at 100K unless the
 crystals fall apart at low T, or if the cryostream is broken.

 I should point out that the first production Cryostream that
 I came across (serial number 2, which I think may have been
 the first one sold!) was in the Cambridge Department of
 Chemistry in about 1985. They didn't become common until the
 mid-1990's in PX labs, when they were already
 well-established as a bit of pretty well essential kit for
 small molecule work.

 So although what Remy says is true, the practice is to
 cryocool most of the time.


 On 19 Jun 2008, at 12:08, Remy Loris wrote:


 Typically crystals of small organic compounds do not
 require freezing as there are no solvent channels. They do in
 general not suffer from radiation damage at room temperature
 the way protein crystals do. Occasionally they are mounted in
 a capillary instead of simply glueing them to a goniometer if
 they are air sensitive. In principle freezing should not
 damage the crystals, but one still may have to be carefull if
 the crystals are large. I think you risk increasing
 mosiacity, and any manipulation that is not needed will on
 average only reduce the quality of the specimen rather than improve
 it

 Remy Loris
 Vrije Univesiteit Brussel

 Jayashankar wrote:

 Dear Scientists and Friends,
 I am not sure, whether  organic crystals  need
 to be in cryo stream necessarily during data  collection from
 an  in house
 xray machine .
 How most of the organic crystals have been
 solved mostly?
 --
 S.Jayashankar
 (A bit confused new generation researcher).
 Research Student
 Institute for Biophysical Chemistry
 Hannover Medical School
 Germany


 Harry
 --
 Dr Harry Powell, MRC Laboratory of Molecular Biology, MRC
 Centre, Hills Road, Cambridge, CB2 2QH







 Disclaimer
 This communication is confidential and may contain privileged
 information intended solely for the named addressee(s). It may not
 be used or disclosed except for the purpose for which it has been
 sent. If you are not the intended recipient you must not review,
 use, disclose, copy, distribute or take any action in reliance upon
 it. If you have received this communication in error, please notify
 Astex Therapeutics Ltd by emailing [EMAIL 

Re: [ccp4bb] is it Ok to freeze

[Santarsiero, Bernard D.]

Dick Dickerson tried to do the same thing at Caltech around the same time.
The major problem with cooling equipment was that the Picker goniometer
had lots of metal in it, and each of the metal pieces cooled and
contracted differently, so the alignment was always off. Nice idea, but
not useful. That's when they got the idea to just cool the sample.

Yes, Sten Samson's device was elegant in a number of ways, and we could
collect 16-20K data for weeks on it.

Bernie Santarsiero

On Thu, June 19, 2008 12:04 pm, Mischa Machius wrote:
 Ha, everyone seems to be bragging about how far back cryo-
 crystallography really goes. In that vain, I'd like to mention that,
 in Martinsried, we had a room that was lined with insulated steel
 walls and that could be flushed with liquid nitrogen. It was requested
 (demanded, really...) by Robert Huber when the Max-Planck Institute
 was finalized in 1972 (I hope I got my history right). That room
 contained an entire diffraction system. Talk about crystal cooling...
 bah, way too dinky. Cool the entire room! Of course, it was a hazard
 to work in that room, and so - as far as I know - there was only one
 post-doc from India how ever used it. That room had an ante-room with
 two more generators plus detectors that could be cooled down to -20°C!
 Ah, the good old Wild West times of macromolecular crystallography...

 Cheers - MM



 On Jun 19, 2008, at 11:48 AM, Pietro Roversi wrote:

 Well everyone, talking of early applications of cryocooling to X-ray
 crystallography, what about Sten Samson's marvellous helium cryostat
 which was operational at Caltech since the end of the 1970s and used
 to
 reach temperatures around 20 K routinely , see for example:

 Proc Natl Acad Sci U S A. 1982 Jul;79(13):4040-4.
 Structure of a B-DNA dodecamer at 16 K.
 Drew HR, Samson S, Dickerson RE.

 That instrument (and its twin) are now both with Riccardo Destro in
 Milano.

  Ciao!

  Pietro




 --
 Pietro Roversi
 Sir William Dunn School of Pathology, Oxford University
 South Parks Road, Oxford OX1 3RE, England UK
 Tel. 0044-1865-275385


 
 Mischa Machius, PhD
 Associate Professor
 Department of Biochemistry
 UT Southwestern Medical Center at Dallas
 5323 Harry Hines Blvd.; ND10.214A
 Dallas, TX 75390-8816; U.S.A.
 Tel: +1 214 645 6381
 Fax: +1 214 645 6353

Re: [ccp4bb] heavy atom derivative(s) for tutorial

[Santarsiero, Bernard D.]

Depending on your time, I'd recommend adding the use of a native gel band
shift to try a one that doesn't work and one that does.

http://www.doe-mbi.ucla.edu/~sawaya/m230d/Crystallization/crystallization.html

I know you can use NaBr or NaI to grow lysozyme crystals as well, and they
stick on hydrophobic patches of the enzyme's surface. I do recall,
however, that the crystal system changes from tetragonal to the triclinic
or monoclinic forms.

Bernie




On Fri, April 11, 2008 8:09 am, Patrick Loll wrote:
 Hi,

 Does anyone have a foolproof recipe for preparing one or more heavy
 atom derivatives for a nice, easily crystallized protein like
 lysozyme?  I'm looking for something that I can use as a hands-on MIR/
 SIR tutorial for beginning students. I'd rather the students be able
 to focus their attention on data collection and analysis, rather than
 spending a lot of time screening different soaks.

 I have looked at Peter Sun's paper on quick soaking (Acta Cryst D58:
 1092-1098, 2002), and will definitely try what they did; but any
 other recipes would also be welcome.

 Thanks,

 Pat


 
 ---
 Patrick J. Loll, Ph. D.
 Professor of Biochemistry  Molecular Biology
 Director, Biochemistry Graduate Program
 Drexel University College of Medicine
 Room 10-102 New College Building
 245 N. 15th St., Mailstop 497
 Philadelphia, PA  19102-1192  USA

 (215) 762-7706
 [EMAIL PROTECTED]

Re: [ccp4bb] Model ensemble for x-ray crystallography

[Santarsiero, Bernard D.]

On Fri, March 28, 2008 10:13 am, Lucas Bleicher wrote:
 Some time ago I've heard about the idea of proposing
 an ensemble of models (as in NMR), instead of a single
 model for x-ray crystallography structures. If I
 remember correctly, this idea has been published
 somewhere. Can anyone tell me what article is that?

 Lucas


   Abra sua conta no Yahoo! Mail, o único sem limite de espaço para
 armazenamento!
 http://br.mail.yahoo.com/



The earliest I recall hearing about this from was W. F. van Gunsteren,
where he used an ensemble average of ten chains instead of a a single atom
chain with anisotropic Gaussian displacement parameters. Check out:

P. Gros, M. Fujinaga, B. W. Dijkstra, K. H. Kalk, and W. G. J. Hol, Acta
Cryst., B45, 488-499 (1989)

Bernie

Re: [ccp4bb] XDS and overlaps

[Santarsiero, Bernard D.]

I use XDSCONV to make the mtz file using all of the reflections. Then I
use uniqueify in CCP4 to make sure the asymmetric unit is correct for
CCP4, and tag the test set.

Bernie Santarsiero


On Thu, February 21, 2008 9:32 am, Dirk Kostrewa wrote:
 Usually, I run CAD first after F2MTZ to make sure that the
 reflections are in the correct reciprocal asymmetric unit for CCP4
 programs. I think, UNIQUE on its own doesn't do this, but the
 UNIQUEIFY script calls CAD, UNIQUE and FREERFLAG for setting a
 FreeR_flag column. The latter may or may not be wanted, depending on
 whether the test-set has been assigned by XDS/XSCALE, already.

 Best regards,

 Dirk.

 Am 21.02.2008 um 16:15 schrieb [EMAIL PROTECTED]:

 In my experience when going from XDS via some intermediate file to
 mtz format, XDS uses in some cases a different reciprocal
 asymmetric unit as mtz uses, which may result in only half of the
 reflections being used and/or ccp4 programs getting confused. By
 using UNIQUE, one makes sure that the reflections are mapped to the
 correct asymmetric unit. It has nothing to do with missing
 reflections but is in many cases essential.

 Best regards,
 Herman


 -Original Message-
 From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf
 Of Kay Diederichs
 Sent: Thursday, February 21, 2008 3:46 PM
 To: CCP4BB@JISCMAIL.AC.UK
 Subject: Re: [ccp4bb] XDS and overlaps

 Simon Kolstoe schrieb:
 Whilst we are on the subject of XDS...

 I had difficulty processing a data-set in mosflm the other day so on
 the recommendation of a colleague switched to xds which, with a
 bit of
 tweaking, seemed to work really well. I converted the resulting
 XDS_ASCII.HKL using xdsconv and then f2mtz ready for phaser and
 refmac.

 We do it in the same way here.


 However, my colleague then told me that xds handled missing
 reflections differently from the usual mosflm/CCP4 route

 I have honestly not the slightest idea what your colleague was
 referring to.

 and thus I had to run the CCP4 program UNIQUE before I tried
 refinement as apparently refmac does not like reflection files
 originally processed with xds. As I couldn't

 In the case of a new project, one should run uniqueify or some
 other means of assigning reflections to the free set (thin shells
 come to mind; see earlier discussions on CCP4BB).

 In the case of an old project, one should transfer the free set of
 reflections from some master data set to the new dataset.

 None of this is specific to XDS.

 HTH,

 Kay

 find anything in the literature about this I was wondering whether
 this advice is still up to date?

 Thanks,

 Simon


 On 21 Feb 2008, at 09:44, Kay Diederichs wrote:

 Engin Ozkan schrieb:
 Hi everyone,
 I have been recently relying on XDS quite a bit, but at the same
 time worrying about how XDS treats overlaps.  We had one dataset
 that both HKL2000 and Mosflm would show to have severe overlaps, as
 expected due to unit cell parameters and the unfortunate crystal
 orientation in the loop. We always ended up with completeness
 percentages in the 70's.
 XDS can find the same lattice, index and scale the data, but yields
 a 100% complete mtz (and a nice structure). Without the
 HKL/Mosflm-like GUI, it is difficult to assess the fate of the
 overlapped observations in XDS. What I could see with VIEW was that
 some observations were being divided into several ovals, probably
 different reflections, but I'm not very certain.
 So, the basic question is, how does XDS treat overlaps?  I could
 not
 find in the documentation an answer to this question; the single
 mention of overlaps I could find tells me that XDS can recognize
 overlaps, but does not tell me if it rejects them, or divvies them
 up into separate reflections, and if that is the case, how does it
 divide them, and how reliable is that? Depending on how it divides
 the overlaps, could that affect commonly-used intensity stats and
 distributions?
 Thanks,
 Engin

 Engin,

 the basic answer is:
 a) each pixel of the detector is assigned to its nearest reflection
 in reciprocal space
 b) some of these pixels will mostly allow the background estimation,
 others will mostly contribute to the integration area (but as they
 are transformed into a local coordinate system there is not a 1:1
 relationship). At this step, pixels which should be background but
 are higher than expected (due to overlap) are rejected.
 c) for each reflection, the background is estimated, and the 3D
 profile is assembled from the pixels contributing to it
 d) a comparison is made: for a reflection, is the percentage of its
 observed profile assembled in c) larger than some constant (called
 MINPK in XDS.INP)? If the answer is no, this reflection will be
 discarded (you could call this situation overlap).

 Among other things, this means that:
 a) the program does _not_ look around each reflection to detect an
 overlap situation, it just tries to gather the pixels for each
 reflection
 b) as a user, when your 

Re: [ccp4bb] differences between Rsym and Rmerge

[Santarsiero, Bernard D.]

I seem to recall hearing Rsym first when I used the Xuong-Hamlin detector,
since there were a substantial number of redundancies. There were two
Rsyms, one called Rrms for the sqrt over the sum of weighted squared
differences and and Rav for the linear summation of unweighted
differences. This was for protein work.

I do recall Rmerge being more popular with the small molecule
crystallographers. However, I also recall a difference between averaging
over pairs of reflections that were or were not Bijvoet pairs, for even
small differences in the anomalous scattering contributions.

Bernie



On Fri, January 18, 2008 9:29 am, R.M. Garavito wrote:
 Kay,

 I beg to differ, but only in a pedantic way.  Historically, Rsym
 would refer to the agreement in symmetry-related reflections within a
 single data set and Rmerge would be the agreement between 2 or more
 data sets that were merged.  This was the way we did it back in the
 old day of precession photography and early oscillation
 photography.  While the terms seem synonymous today, the recent
 thread [ccp4bb] combine incomplete data sets illustrates where such
 a distinction is still relevant, where the merging is between data
 collected under different experimental conditions (i.e., a different
 crystal in a different orientation).

 Michael

 
 R. Michael Garavito, Ph.D.
 Professor of Biochemistry  Molecular Biology
 513 Biochemistry Bldg.
 Michigan State University
 East Lansing, MI 48824-1319
 Office:  (517) 355-9724 Lab:  (517) 353-9125
 FAX:  (517) 353-9334Email:  [EMAIL PROTECTED]
 


 On Jan 18, 2008, at 10:15 AM, Kay Diederichs wrote:

 Salameh, Mohd A., Ph.D. schrieb:
 Hi everybody!
 I will appreciate it if anybody can clarify to me the differences
 between Rmerge and Rsym. Many thanks, M

 there is no difference - unfortunately there are two words for the
 same thing. Rmerge currently appears to be more in fashion.

 just my 2 cents,

 Kay
 --
 Kay Diederichshttp://strucbio.biologie.uni-konstanz.de
 email: [EMAIL PROTECTED]Tel +49 7531 88 4049 Fax 3183
 Fachbereich Biologie, Universität Konstanz, Box M647, D-78457 Konstanz

Re: [ccp4bb] differences between Rsym and Rmerge

[Santarsiero, Bernard D.]

You know there is that other funny column with chi^2's. I like to quote
both. Half of the reviewers will know which column to look at, but you
will satisfy the other half.

Bernie

On Fri, January 18, 2008 1:39 pm, Edwin Pozharski wrote:
 There are two opposing views on this.

 First:  Rmerge doesn't matter.  Don't even look into that column in
 scalepack output, you will be upset over nothing.  If you collect twice
 as much data (360 sweep instead of 180) from the same crystal, your
 Rmerge will go up due to higher redundancy, but the dataset will
 actually get better because you measuring every reflection twice more
 and your I/sigma will increase by ~40%.

 Second:  Rmerge is very important, because if it is, say, 100% (oh,
 those zeros in the scalepack output) it means that symmetry-related
 reflections vary by about 100%, so your data is a pile of garbage (at
 least in that resolution shell).  Cut your data at the resolution where
 Rmerge is 30% and you will be rewarded by really low Rfactors for your
 final model.  Plus, if you keep all the data to where I/sigma~1, your
 Rmerge is guaranteed to be 0.00 in the output, and what are you going to
 tell reviewers of your paper?

 Of course, truth is somewhere in the middle.  If I collect on two
 crystals of the same type (assuming everything else is the same, such as
 redundancy), and one has much higher Rmerge, then I should probably
 choose the other one.  If you cut resolution at I/sigma~1, and your
 overall Rmerge is about 10%, I think it's normal.  But if it's 30%, you
 may have some unusually high level of noise in your data (satellite
 crystal?  twinning?  evil xray fairy messing with you?).  So Rmerge does
 tell you something, but only in context with all the other information.
 After all, the only thing that matters is if your electron density map
 is interpretable.  I dare to say that the quality of the map you get
 does correlate with Rmerge, but would I discard a dataset just because
 Rmerge is high without trying to solve the structure and take a look at
 the density?  Never.

 Cheers,

 Ed.

 Mischa Machius wrote:
 OK, that brings us back to a more substantial question: is any of
 these R values actually suitable to judge the quality of a given
 dataset? Instead of introducing novel R factors, one could also simply
 ignore them altogether, make sure that the error models have been
 properly chosen and look at I/sigma(I) as the main criterion. [QUOTE
 ]If anyone then still wants to present low R factors, one can always
 divide by 2, if necessary. [/QUOTE]

 Best - MM


 On Jan 18, 2008, at 1:02 PM, Salameh, Mohd A., Ph.D. wrote:

 Thank you all, it was very, very helpful discussion. However, I
 collected crystal data and the Rmerge overall was very high around 0.17
 at 2.6A resolution and I'm wondering what is the acceptable value
 (range) of R-merge that worth the time to continue processing! Very
 anxious to hear your thoughts. Thanks, M
 
 Mohammed A. Salameh, Ph.D.
 Mayo Clinic Cancer Center
 Griffin Cancer Research Building
 4500 San Pablo Road
 Jacksonville, FL 32224
 Tel:(904) 953-0046
 Fax:(904) 953-0277
 [EMAIL PROTECTED]
 


 -Original Message-
 From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of
 Chris Putnam
 Sent: Friday, January 18, 2008 1:21 PM
 To: CCP4BB@JISCMAIL.AC.UK
 Subject: Re: [ccp4bb] differences between Rsym and Rmerge

 On Friday 18 January 2008 09:30:06 am Ethan A Merritt wrote:

 Rmerge is an average over replicate measurements of the intensity for
 identical [hkl]. Rsym is an average over the measurements for all
 symmetry
 equivalent reflections.

 In the presence of anomalous scattering, Rsym will be higher than
 Rmerge
 because the Bijvoet pairs, although symmetry related, do not have
 identical
 intensities.

 One might logically report two values for Rsym,  one which averages
 over the Bijvoet-paired reflections and one which does not.


 This has been an eye-opening discussion for me.  I've been really
 surprised
 that there's been such a diversity of opinion about what these common
 terms ought to refer to, and the fact that my understanding was wrong.
 I always thought that Rsym was an average over all symmetry equivalent
 reflections from the same crystal (including Bijvoet pairs) and Rmerge
 was
 properly restricted to cases of multi-crystal averaging.  (My versions
 of
 Table 1's from single crystals have used Rsym rather than Rmerge.)

 I wonder if the problem here is that the terms have become overloaded
 (and
 hence non-specific).  In that sense Rmerge is a particularly
 unfortunate
 name as every R that we're discussing is a really a merge of some sort
 or
 another.  (In the most naive sense, Rmerge might be thought to be the
 R
 for whatever variation of reflection merging the experimenter chooses
 to
 do.)

 One possible solution would be to push the community towards a new 

Re: [ccp4bb] differences between Rsym and Rmerge

[Santarsiero, Bernard D.]

Ed,

You don't need to adjust them in XDS.

In scalepack, I don't adjust the individual chi^2's (the additive terms,
by shell), but I do adjust the multiplier to get the chi^2's in the
highest resolution shell, usually with I/sig (or I/sig) ~ 1, to be
near and greater than 1.0. Then the overall chi^2's are what they are. My
rationale is that the weakest shell will be dominated most by random
errors, and should approach 1.0, whereas the lower resolution shells
typically have chi^2's greater than 1, and are dominated by systematic
(absorption, misalignment, integration) errors. If they aren't close to 1,
then that's telling you something, whether or not the R(sym) is.

I report the overall chi^2 and the value in the highest resolution shell,
and they're in the pdb/cif file.

Bernie

On Fri, January 18, 2008 2:28 pm, Edwin Pozharski wrote:
 Bernie,

 but my chi-squares are always near 1.0, so why would I report it?  How
 close they should be to 1 is open to discussion, of course.   The point
 is, it is assumed (at least in scalepack) that you adjust your error
 model until chi-square~1.  I have never seen a statistics table in a
 paper which would report chi-suqares.

 I am afraid I may be misinterpreting what you were trying to say - I
 apologize if that is the case.

 Cheers,

 Ed.

 Santarsiero, Bernard D. wrote:
 You know there is that other funny column with chi^2's. I like to quote
 both. Half of the reviewers will know which column to look at, but you
 will satisfy the other half.

 Bernie

 On Fri, January 18, 2008 1:39 pm, Edwin Pozharski wrote:

 There are two opposing views on this.

 First:  Rmerge doesn't matter.  Don't even look into that column in
 scalepack output, you will be upset over nothing.  If you collect twice
 as much data (360 sweep instead of 180) from the same crystal, your
 Rmerge will go up due to higher redundancy, but the dataset will
 actually get better because you measuring every reflection twice more
 and your I/sigma will increase by ~40%.

 Second:  Rmerge is very important, because if it is, say, 100% (oh,
 those zeros in the scalepack output) it means that symmetry-related
 reflections vary by about 100%, so your data is a pile of garbage (at
 least in that resolution shell).  Cut your data at the resolution where
 Rmerge is 30% and you will be rewarded by really low Rfactors for your
 final model.  Plus, if you keep all the data to where I/sigma~1, your
 Rmerge is guaranteed to be 0.00 in the output, and what are you going
 to
 tell reviewers of your paper?

 Of course, truth is somewhere in the middle.  If I collect on two
 crystals of the same type (assuming everything else is the same, such
 as
 redundancy), and one has much higher Rmerge, then I should probably
 choose the other one.  If you cut resolution at I/sigma~1, and your
 overall Rmerge is about 10%, I think it's normal.  But if it's 30%, you
 may have some unusually high level of noise in your data (satellite
 crystal?  twinning?  evil xray fairy messing with you?).  So Rmerge
 does
 tell you something, but only in context with all the other information.
 After all, the only thing that matters is if your electron density map
 is interpretable.  I dare to say that the quality of the map you get
 does correlate with Rmerge, but would I discard a dataset just because
 Rmerge is high without trying to solve the structure and take a look at
 the density?  Never.

 Cheers,

 Ed.

 Mischa Machius wrote:

 OK, that brings us back to a more substantial question: is any of
 these R values actually suitable to judge the quality of a given
 dataset? Instead of introducing novel R factors, one could also simply
 ignore them altogether, make sure that the error models have been
 properly chosen and look at I/sigma(I) as the main criterion. [QUOTE
 ]If anyone then still wants to present low R factors, one can always
 divide by 2, if necessary. [/QUOTE]

 Best - MM


 On Jan 18, 2008, at 1:02 PM, Salameh, Mohd A., Ph.D. wrote:


 Thank you all, it was very, very helpful discussion. However, I
 collected crystal data and the Rmerge overall was very high around
 0.17
 at 2.6A resolution and I'm wondering what is the acceptable value
 (range) of R-merge that worth the time to continue processing! Very
 anxious to hear your thoughts. Thanks, M
 
 Mohammed A. Salameh, Ph.D.
 Mayo Clinic Cancer Center
 Griffin Cancer Research Building
 4500 San Pablo Road
 Jacksonville, FL 32224
 Tel:(904) 953-0046
 Fax:(904) 953-0277
 [EMAIL PROTECTED]
 


 -Original Message-
 From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of
 Chris Putnam
 Sent: Friday, January 18, 2008 1:21 PM
 To: CCP4BB@JISCMAIL.AC.UK
 Subject: Re: [ccp4bb] differences between Rsym and Rmerge

 On Friday 18 January 2008 09:30:06 am Ethan A Merritt wrote:

 Rmerge is an average over replicate measurements of the intensity
 for
 identical [hkl

Re: [ccp4bb] crystallisation robot

[Santarsiero, Bernard D.]

In our original work on a prototype, we used the Cartesian technology. We
were able to dispense 10nL+10nL drops with a large range of viscosities
without difficulty. The main issue was to wash the tips after use to
prevent clogging. That was with a system that could dispense 1nL droplets.
The present systems are sub-nL now. We didn't have a problem with
electrostatics.

More recently, I've looked at all of the crystallization robot vendors.
For single lab users, all of the systems work well. Systems like the Hydra
or Mosquito are less automatic, but provide the basic functions for
crystallization trial setup. For more of a user facility with a large
number of users from several groups, you want more automation to avoid
protocols that can damage the components, like alignment or breakage of
the needles. You also want to consider the total annual cost of
expendables and maintenance.

Bernie Santarsiero

On Wed, January 16, 2008 5:46 am, Demetres D. Leonidas wrote:
 Oryxnano 50+50 nL

 Demetres


 David Briggs wrote:
 I'll defend the honour of the phoenix... (again)
 Bernhard Rupp 100+100 nl
 Dave Briggs (and all users at Univ of Manchester, UK) 100+100nl Others..
 Only time we have ANY problems is when the nano dispensing tip gets
clogged. Often a good wash whilst still on the machine will clear the
blockage.
 Dave
 --
 
 David C. Briggs PhD
 Father  Crystallographer
 http://www.dbriggs.talktalk.net http://www.dbriggs.talktalk.net AIM
ID: dbassophile
 

 --
 Demetres D. Leonidas, Ph.D.
 Structural Biology  Chemistry Group
 Institute of Organic and Pharmaceutical Chemistry
 The National Hellenic Research Foundation
 48, Vassileos Constantinou Avenue
 Athens 116 35, Greece
 ==
 Tel. +30 210 7273841 (office)
  +30 210 7273895 (lab)
 Fax. +30 210 7273831
 E-mail: [EMAIL PROTECTED]
 URL: http://athena.eie.gr
 ==

Re: [ccp4bb] Fwd: [ccp4bb] Occupancy refinement of substrate using refmac5

[Santarsiero, Bernard D.]

I would also ask what is the actual goal in refining the occupancy of the
ligand?

While I agree wholeheartedly with George, the B-factors will adequately
model a lower ligand occupancy.  Usually the question you want to resolve
is Is the ligand really bound in the active site, or is this an artifact
of the model?  However, if you're trying to resolve a static disorder in
an active site, say half ligand and half solvent occupancies, then
refining the ligand occupancy can be useful.

Thus, what you really need to do is compare a refinement with just
B-factors and full occupancies vs. a refinement with the added restrained
ligand occupancy variable, in some sort of R-test, comparison of GOF, etc.
Does it really improve the model?

Bernie Santarsiero

On Mon, December 17, 2007 8:43 am, George M. Sheldrick wrote:
 I thought that I would never have to disagree with both Eleanor and
 Tassos in the same email, let alone risk being burnt at a stake as a
 heretic for doubting the Gospel according to Kleywegt, but in my
 experience, given the very fortunate position that you have data to
 1.5A, the refinement of one occupancy parameter for the whole ligand
 (e.g. one SHELX free variable) will be rather well defined provided
 that sensible restraints are applied to the B-values. A common variant
 of this for ligands or side-chains is to refine one component with
 occupancy p and an alternative component or conformation with an
 occupancy 1-p, still only requiring the addition of ONE refinable
 parameter. If you are using SHELX and your ligand (or part of it)
 happens to be rigid, the rigid group refinement offers a further
 way of keeping the number of refinable parameters down. The Dundee
 PRODRG server is an excellent source of ligand geometries and
 restraints for such refinements.

 George

 Prof. George M. Sheldrick FRS
 Dept. Structural Chemistry,
 University of Goettingen,
 Tammannstr. 4,
 D37077 Goettingen, Germany
 Tel. +49-551-39-3021 or -3068
 Fax. +49-551-39-2582


 On Mon, 17 Dec 2007, Anastassis Perrakis wrote:

  I have already changed occupancies as Eleanor mentioned, and got
  approximate values. But my hope is to try to get much precise ones if
  possible.
 
 I never expected to preach the 'Kleywegt Gospel' in the ccp4bb,
 but in this case what you need is more accurate answers, not more
 precise ones
 (or better both, but precision alone can be a problem, and you can
 easily get
 'precise' but inaccurate data easily by making the wrong assumptions
 in your experiment)

 http://en.wikipedia.org/wiki/Accuracy

  I have heard from my colleague SHELX can refine occupancies, and
  got its license. I'll next try SHELX.

 I think that phenix.refine can also do occupancies ?
 The problem is not  if the program can do it, but if at your specific
 case
 you have enough information to do that in a meaningful way.

 For a soaking experiment and 1.5 A data, I would say that Eleanor's
 suggestion
 of tuning Occ based on B, is as close as you would get, accurate enough
 given
 the data,
 although not necessarily too precise.

 Tassos

Re: [ccp4bb] To bathe or not to bathe.

[Santarsiero, Bernard D.]

The main reason was related to absorption. If you didn't completely bathe
the crystal in the xray beam, then the diffracting volume of the crystal
would be different during the data collection, and thus, scaling would be
inaccurate, especially when there was radiation damage. This was
especially true when you weren't sure that the crystal was well-centered
in the xray beam (in a cryostat, and therefore not visible). We typically
collected highly redundant data to help compensate for this. We also used
to correct for absorption by assigning Bragg indices to the crystal and
making precise measurements of crystal dimensions.

Scaling programs are now more extensive, and include options to calculate
a pseudo-absorption surface. In principle, if you have a beam that is
ALWAYS smaller than the crystal, then the same crystal volume is
illuminated by the xray beam, and will minimize scaling errors.

Bernie Santarsiero



On Fri, November 23, 2007 4:34 pm, Jim Pflugrath wrote:
 It probably goes back to the days of using a single-counter diffractometer
 where one didn't have multiple Bragg reflections on an image or film pack.
 That is, each reflection was collected by itself.  Even in a small
 molecule
 crystal data collection nowadays, it would not hurt to have the crystal
 completely bathed in the beam.

 Also in the old days (let's say pre-cryo), there was plenty of radiation
 damage going on even with a sealed-tube source.  We always corrected for
 radiation damage by extrapolating back to zero dose in those days.

 Jim

 -Original Message-
 From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of
 Robert
 Sweet
 Sent: Friday, November 23, 2007 4:08 PM
 To: CCP4BB@JISCMAIL.AC.UK
 Subject: [ccp4bb] To bathe or not to bathe.

 Jorge,

 You said,

   I remember one former good (small molecule ?) crystallography book
 with words a kind of this the crystals should be completely bathed by
 the x-ray beam during the whole data collection  ...

 The original motive for bathing the whole crystal was to assure that the
 relative intensity of the data on each successive film pack was very
 nearly constant.  This was possible (one might say necessary) in the old
 days because the laboratory sources were very stable and the intensity was
 low enough that there wasn't a lot of x-ray damage to the crystals.
 There were a couple of other good reasons to pay attention to details like
 this.  One was that methods for scaling images together were not quite as
 good as now, and another was that film data were relatively very much less
 accurate than what is achievable now with excellent detectors and brighter
 sources.  To combat all of that, we tried to do everything possible to
 make things better.

 These days scaling algorithms are good, the detectors are excellent, and
 very often it pays to employ a beam smaller than the x-tal.  This, the
 non-uniformity of many synchrotron beams, and the systematic damage
 to crystals that we observe now with synchrotron sources cause serious
 systematic errors.  We're forced to depend on good scaling and good
 detectors to get accurate measurements.  Making the measurements in many
 different crystal orientations (redundancy) helps to smooth out these
 systematic errors.

 Nonetheless, it will always pay you to watch for EACH of these sources of
 error and to minimize them as best you can.

 Bob

 =
  Robert M. Sweet E-Dress: [EMAIL PROTECTED]
  Group Leader, PXRR: Macromolecular   ^ (that's L
Crystallography Research Resource at NSLSnot 1)
http://px.nsls.bnl.gov/
  Biology Dept
  Brookhaven Nat'l Lab.   Phones:
  Upton, NY  11973631 344 3401  (Office)
  U.S.A.  631 344 2741  (Facsimile)
 =

Re: [ccp4bb] difference of R_work and R_free as a function of resolution

[Santarsiero, Bernard D.]

One of the major problems with any R-factor is that it's a function of the
denominator, and therefore I, F^2, or F. Depending on how you push to
higher resolution, the R's will very likely increase due to the dominance
of the denominator over the numerator. It's good to also monitor on
evaluation functions that follow the numerator, like chi^2 in
scaling/averaging, and the figure-of-merit (rho-calc vs. rho-obs), and
goodness-of-fit.

Bernie Santarsiero

On Thu, November 15, 2007 10:41 am, Kay Diederichs wrote:
 Dear crystallographers,

 today I explained to a student that I believe that the difference of
 R_work and R_free should decrease as a function of resolution, because
 at high resolution there is less danger of overfitting, whereas at low
 resolution one is always overfitting to some extent. I'd say this should
 be true at least in absolute terms, but probably even in relative terms:
 for example, for a 1.5A structure I'd expect R_work/R_free= 14%/16%,
 whereas at 3A I'd expect 28%/35%.

 I believe that I saw plots (maybe in a link given in a posting on
 CCP4BB) obtained for structures from the PDB, which confirmed this
 hypothesis. But I've been googling and searching through postings now
 for quite some while, and cannot seem to find anything to that effect.

 Does anybody have a pointer to an analysis of this effect?

 thanks,

 Kay

 P.S. Would be good to put this into the CCP4 Wiki article on R-factors
 --
 Kay Diederichshttp://strucbio.biologie.uni-konstanz.de
 email: [EMAIL PROTECTED]Tel +49 7531 88 4049 Fax 3183
 Fachbereich Biologie, Universität Konstanz, Box M647, D-78457 Konstanz

Re: [ccp4bb] Wavelength of Copper Kalpha

[Santarsiero, Bernard D.]

On Wed, October 10, 2007 11:50 am, Bryan W. Lepore wrote:
 On Wed, 10 Oct 2007, Jim Pflugrath wrote:
 It has come to my attention that the wavelength of a Copper Kalpha may
 have changed over the years.  At least this appears to be true if you
 look at the International Tables.

 the 'natural' isotopic distribution must have changed, as it has with the
 biggest example lithium, since the 'nuclear age' - i take that to be the
 principal reason.

 and i take it that everyday-use copper anodes are isotopically impure?

 -bryan

The change was probably around 1967, when the meter was defined, slightly,
using krypton. Tungsten was used as a reference, and the kX to A ratio was
changed: Bearden, Rev Mod Phys, 39, 78 (1967).

old values from IT, volume III
alpha-1 1.54051
alpha-2 1.54433
beta-1  1.39217
beta-2  1.38102

alpha   1.54178

the ratio of I(alpha-2)/I(alpha-1) = 0.497
the units are kX/1.00202, and good to 1 in 25,000, based on kX, for 1962
(IT, 1962).

newer values, from IT, volume IV

alpha-1 1.540562
alpha-2 1.544390
beta-1  1.392218
beta-5  1.38109

based on the W K-alpha-1 line/peak at 0.2090100 A

Bernie Santarsiero

Re: [ccp4bb] Popularity of Stereo graphics in crystallography

[Santarsiero, Bernard D.]

I agree with Kevin. We have stereo on about half of our workstations, and
no one has used them in about three years. We typically use O.

Also, we have three large servers which are relatively fast. So the main
purpose of a workstation is building, not computing here. That way you can
easily work on multiple structures on a workstation at the same time,
while you're refining and building them. We have a few people that use
PC's and Coot as well.

Bernie


On Wed, June 20, 2007 11:45 am, P Hubbard wrote:
 Hi,

 Thanks for the e-mail. The current results of the survey would certainly
 put
 you in the minority! Stereo graphics are not dead after all.

 I have used systems with and without stereo graphics. I personally prefer
 them, and think they are great for helping newbies refine, and for
 non-structural biologists and students to look at molecular architecture.
 It
 seems a lot of other people, for whatever reason, like them too.

 Paul


From: Kevin Cowtan [EMAIL PROTECTED]
Reply-To: Kevin Cowtan [EMAIL PROTECTED]
To: CCP4BB@JISCMAIL.AC.UK
Subject: Re: [ccp4bb] Survey on computer usage in crystallography
Date: Wed, 20 Jun 2007 17:27:04 +0100

More likely the issue is that some of us do not find stereo to be
 necessary
of beneficial for crystallographic model building.

In which case, given the power of modern PCs and graphics cards, a basic
off-the-shelf PC costing $1000/£500 is completely adaquate for typical
structure solution and model building problems.

I use coot a lot and I haven't even bothered installing the graphics
drivers for my graphics card. All the 3D stuff gets gone in software, and
most of the graphics hardware sits around doing nothing. If I needed the
performance, it would be a 5 minute job to install the drivers, but I
haven't needed it.

Kevin

P Hubbard wrote:
I am sorry you are unhappy with the questions, David.

As I am sure you know, I half-decent system with stereo graphics doesn't
come cheap, and if you price things together to make something that
performs well I doubt you'll get much change out of $2000.

I am aware of other 3D systems (such as those listed on
 www.stereo3d.com).
However, the price of peripherals like a 3D LCD monitor are
 prohibitively
expensive (and the quality of the images is supposed to be poor). Do you
know of a relatively inexpensive way of displaying 3D images on PCs?

Any other comments would be greatly appreciated.

Paul

 _
 Get a preview of Live Earth, the hottest event this summer - only on MSN
 http://liveearth.msn.com?source=msntaglineliveearthhm

Re: [ccp4bb] Highest shell standards

[Santarsiero, Bernard D.]

I seem to recall this being discussed at some point.

For the difference electron density map, there clearly isn't a downside to
loss of reflections, i.e., the coefficients in the map generation are
formally zero for Fo-Fc (which all the scaling, weight, sigma-A bits in
there). If the phases are fairly good, then you are assuming that the Fobs
agrees perfectly with the Fcalc. You don't gain any new details in the
map, but the map isn't distorted with loss of these terms.

As for the 2Fo-Fc (2mFo-DFc, or something like that) electron density map,
it again assumes that the phases are in good shape, and you essentially
lose any new information you could gain from the addition of new Fobs
terms, but the map isn't distorted since the terms are zero.

I seem to recall in the dark ages that you could make an Fobs map, and
included Fcalc, or some fraction of Fcalc (like 0.5Fc) in as the Fobs term
for missing reflections. That way the amplitudes are closer to being
correct for resolution shells that are fairly incomplete. Anyone else
remember this for small molecule structures?

The real issue, generally, is that the phases are the most important
factor in making a good map, and the structure factors are, unfortunately,
weaker contributors to features in the maps.

Bernie




On Fri, March 23, 2007 4:02 am, Eleanor Dodson wrote:
 This is a good point - I had thought that D would be very low for an
 incomplete shell, but that doesnt seem to be true..

 Garib - what do you think?
 Eleanor


 Petrus H Zwart wrote:
 I typically process my data to a maximum I/sig near 1, and
 completeness in
 the highest resolution shell to 50% or greater. It


 What about maps computed of very incomplete datasets at high resolution?
 Don't you get a false sense of details when the missing reflections are
 filled in with DFc when computing a 2MFo-DFc map?

 P

Re: [ccp4bb] Highest shell standards

[Santarsiero, Bernard D.]

There are journals that have specific specifications for these parameters,
so it matters where you publish. I've seen restrictions that the highest
resolution shell has to have I/sig  2 and completeness  90%. Your
mileage may vary.

I typically process my data to a maximum I/sig near 1, and completeness in
the highest resolution shell to 50% or greater. It's reasonable to expect
the multiplicity/redundancy to be greater than 2, though that is difficult
with the lower symmetry space groups in triclinc and monoclinic systems
(depending upon crystal orientation and detector geometry). The chi^2's
should be relatively uniform over the entire resolution range, near 1 in
the highest resolution bins, and near 1 overall. With this set of
criteria, R(merge)/R(sym) (on I) can be as high as 20% and near 100% for
the highest resolution shell. R is a poor descriptor when you have a
substantial number of weak intensities because it is dominated by the
denominator; chi^2's are a better descriptor since it has, essentially,
the same numerator.

One should also note that the I/sig criteria can be misleading. It is the
*average* of the I/sig in a resolution shell, and as such, will include
intensities that are both weaker and stronger than the average. For the
highest resolution shell, if you discard intensities greater than 2sig,
then you are also discarding intensities substantially greater than 2sig
as well. The natural falloff of the intensities is reflected (no pun
intended) by the average B-factor of the structure, and you need the
higher resolution, weaker data to best define that parameter.

Protein diffraction data is inherently weak, and far weaker than we obtain
for small molecule crystals. Generally, we need all the data we can get,
and the dynamic range of the data that we do get is smaller than that
observed for small molecule crystals. That's why we use restraints in
refinement. An observation of a weak intensity is just as valid as the
observation of a strong observation, since you are minimizing a function
related to matching Iobs to Icalc. This is even more valid with refinement
targets like the maximum likelihood function. The ONLY reasons that we
ever used I/sig or F/sig cutoffs in refinements was to make the
calculations faster (since we were substantially limited by computing
power decades ago), the sig's were not well-defined for weak intensities
(especially for F's), and the detectors were not as sensitive. Now, with
high brilliance x-ray sources and modern detectors, you can, in fact,
measure weak intensities well--far better than we could decades ago. And
while the dynamic range of intensities for a protein set is relatively
flat, in comparison to a small molecule dataset, those weak terms near
zero are important in restraining the Fcalc's to be small, and therefore
helping to define the phases properly.

In 2007, I don't see a valid argument of severe cutoff's in I/sig at the
processing stage. I/sig = 1 and a reasonable completeness of 30-50% in the
highest resolution shell should be adequate to include most of the useful
data. Later on, during refinement, you can, indeed, increase the
resolution limit, if you wish. Again, with targets like maximum
likelihood, there is no statistical reason to do that. You do it because
it makes the R(cryst), R(free), and FOM look better. You do it because you
want to have a 2.00A vs. 1.96A resolution structure. What is always true
is that you need to look at the maps, and they need as many terms in the
Fourier summation as you can include. There should never be an argument
that you're savings on computing cycles. It's takes far longer to look
carefully at an electron density map and make decisions on what to do than
to carry out refinement. We're rarely talking about twice the computing
time, we're probably thinking 10% more. That's definately not a reason to
throw out data. We've got lots of computing power and lots of disk
storage, let's use to our advantage.

That's my nickel.

Bernie Santarsiero




On Thu, March 22, 2007 7:00 am, Ranvir Singh wrote:
 I will agree with Ulrich. Even at 3.0 A, it is
 possible to have a  structure with reasonable accuracy
 which can explain the biological function/ or is
 consistent with available biochemical data.
 Ranvir
 --- Ulrich Genick [EMAIL PROTECTED] wrote:

 Here are my 2-3 cents worth on the topic:

 The first thing to keep in mind is that the goal of
 a structure
 determination
 is not to get the best stats or to claim the highest
 possible
 resolution.
 The goal is to get the best possible structure and
 to be confident that
 observed features in a structure are real and not
 the result of noise.

  From that perspective, if any of the conclusions
 one draws from a
 structure
 change depending on whether one includes data with
 an I/sigI in the
 highest
 resolution shell of 2 or 1, one probably treads on
 thin ice.

 The general guide that one should include only data,
 for which the
 shell's average
   I/sigI  2 

Re: [ccp4bb] Highest shell standards

[Santarsiero, Bernard D.]

My guess is that the integration is roughly the same, unless the profiles
are really poorly defined, but that the scaling that is suffering from
using a lot of high-resolution weak data. We've integrated data to say
I/sig = 0.5, and sometimes seem more problems with scaling. I then cut
back to I/sig = 1 and it's fine. The major difficulty arises that if the
crystal is dying, and the decay/scaling/absorption model isn't good
enough. So that's definately a consideration when trying to get a more
complete data set and higher resolution (so more redundancy).

Bernie


On Thu, March 22, 2007 12:21 pm, Jose Antonio Cuesta-Seijo wrote:
 I have observed something similar myself using Saint in a Bruker
 Smart6K detector and using denzo in lab and syncrotron detectors.
 First the I over sigma never really drops to zero, no mater how much
 over your real resolution limit you integrate.
 Second, if I integrate to the visual resolution limit of, say, 1.5A,
 I get nice dataset statistics. If I now re-integrate (and re-scale)
 to 1.2A, thus including mostly empty (background) pixels everywhere,
 then cut the dataset after scaling to the same 1.5A limit, the
 statistics are much worse, booth in I over sigma and Rint. (Sorry, no
 numbers here, I tried this sometime ago).
 I guess the integration is suffering at profile fitting level while
 the scaling suffers from general noise (those weak reflections
 between 1.5A and 1.2A will be half of your total data!).
 I would be careful to go much over the visual resolution limit.
 Jose.

 **
 Jose Antonio Cuesta-Seijo
 Cancer Genomics and Proteomics
 Ontario Cancer Institute, UHN
 MaRs TMDT Room 4-902M
 101 College Street
 M5G 1L7 Toronto, On, Canada
 Phone:  (416)581-7544
 Fax: (416)581-7562
 email: [EMAIL PROTECTED]
 **


 On Mar 22, 2007, at 10:59 AM, Sue Roberts wrote:

 I have a question about how the experimental sigmas are affected
 when one includes resolution shells containing mostly unobserved
 reflections.  Does this vary with the data reduction software being
 used?

 One thing I've noticed when scaling data (this with d*trek (Crystal
 Clear) since it's the program I use most) is that I/sigma(I) of
 reflections can change significantly when one changes the high
 resolution cutoff.

 If I set the detector so that the edge is about where I stop seeing
 reflections and integrate to the corner of the detector, I'll get a
 dataset where I/sigma(I) is really compressed - there is a lot of
 high resolution data with I/sigma(I) about 1, but for the lowest
 resolution shell, the overall I/sigma(I) will be maybe 8-9.  If the
 data set is cutoff at a lower resolution (where I/sigma(I) in the
 shell is about 2) and scaled, I/sigma(I) in the lowest resolution
 shell will be maybe 20 or even higher (OK, there is a different
 resolution cutoff for this shell, but if I look at individual
 reflections, the trend holds).  Since the maximum likelihood
 refinements use sigmas for weighting this must affect the
 refinement.  My experience is that interpretation of the maps is
 easier when the cut-off datasets are used. (Refinement is via
 refmac5 or shelx).  Also, I'm mostly talking about datasets from
 well-diffracting crystals (better than 2 A).

 Sue


 On Mar 22, 2007, at 2:29 AM, Eleanor Dodson wrote:

 I feel that is rather severe for ML refinement - sometimes for
 instance it helps to use all the data from the images, integrating
 right into the corners, thus getting a very incomplete set for the
 highest resolution shell.  But for exptl phasing it does not help
 to have many many weak reflections..

 Is there any way of testing this though? Only way I can think of
 to refine against a poorer set with varying protocols, then
 improve crystals/data and see which protocol for the poorer data
 gave the best agreement for the model comparison?

 And even that is not decisive - presumably the data would have
 come from different crystals with maybe small diffs between the
 models..
 Eleanor



 Shane Atwell wrote:

 Could someone point me to some standards for data quality,
 especially for publishing structures? I'm wondering in particular
 about highest shell completeness, multiplicity, sigma and Rmerge.

 A co-worker pointed me to a '97 article by Kleywegt and Jones:

 _http://xray.bmc.uu.se/gerard/gmrp/gmrp.html_

 To decide at which shell to cut off the resolution, we nowadays
 tend to use the following criteria for the highest shell:
 completeness  80 %, multiplicity  2, more than 60 % of the
 reflections with I  3 sigma(I), and Rmerge  40 %. In our
 opinion, it is better to have a good 1.8 Å structure, than a poor
 1.637 Å structure.

 Are these recommendations still valid with maximum likelihood
 methods? We tend to use more data, especially in terms of the
 Rmerge and sigma cuttoff.

 Thanks in advance,

 *Shane Atwell*


 Sue Roberts
 Biochemistry  Biopphysics
 University of Arizona

 [EMAIL PROTECTED]

[ccp4bb] XDS to MTZ for SOLVE

[Santarsiero, Bernard D.]

Not purely a ccp4 question, but an MTZ file is involved, so stick with me.

I've collected a SAD data set, and processed with XDS. I can use XDSCONV
to generate an MTZ, SHELX, and CNS file. I chose the MTZ file since it
keeps the intensities. SHELX does too (the F**2), but has separate lines
for (h,k,l) and (-h,-k,-l). The MTZ file has hkl,IMEAN, SIGIMEAN, I+,
SIGI+, I-, SIGI-.

I used mtzdump (or could use MTZVARIOUS) to convert this file to a simple
ASCII file, and then extracted just the hkl, I+, SIGI+, I-, SIGI-. Since
CCP4 fills in all of the reflections in an ASU, I removed those reflection
entries with SIGIMEAN = -999.0.

I have a file with has some entries with -999 (for the non-measured
value), especially those in the centric zones.

For SOLVE, does it identify entries with a value of -999 as a placeholder,
or should I entirely removed those from the data file? I used the options

PREMERGED
READFORMATTED
READ_INTENSITIES

in SOLVE. What if I measured I- but I+? That's why I left them all in.

Suggestions on what is best for the solve run? I'm concerned that it's
interpretting the -999 as a real measurement.

Bernie Santarsiero

Re: [ccp4bb] practical limits of MR?

[Santarsiero, Bernard D.]

On Mon, March 5, 2007 2:16 pm, Nat Echols wrote:
 I had a debate with a coworker about using MR in desperation and I'm
 curious what the most extreme case is where a very different model was
 used to solve a structure.  This could be highest RMSD, lowest % identity,
 or most incomplete model.  I'm also curious whether homology modelling has
 ever been useful for this.  (I'm pretty sure I've come across papers
 discussing this last concept.)

 thanks,
 Nat


I'm not sure what in desperation means. You're trying to solve a
structure, so all options are open, right? They don't have to be elegant,
it just has to work. :)

I know of a few cases where roughly 15% of a structure was used as a
search model to solve a structure. Recently I used about 30% of the
structure to get the initial model with MR, and then bootstrap along to
find the remainder. Similarly, you can get away with 25-30% sequence
homology (so long as there is high structural homology) with MR. It's
relatively easy to find a helix-turn-helix motif by MR than, say, an
extended beta-sheet expanse, because the beta-sheet can substantially bend
and twist, whereas the helices are fairly rigid.

Bernie Santarsiero

Re: [ccp4bb] how many stuck datasets are actually twinned?

[Santarsiero, Bernard D.]

It's very possible with any overlap of lattice points, even in the lower
symmetry space groups. To extend Eleanor's list: For example, I once had a
structure with the unit cell relationship 3c cos(beta) = -a. In cases like
that, it's not really a clean diffraction, but looks very much like a
single lattice of points. This is so-called reticular pseudomerohedry
across the ab-plane with a twin index of 3. The h,k,l reflections of the
parent lattice are superimposed upon h, k, (-l-2h)/3 reflections of the
twin lattice. So twinning can be frightenly common.

On Wed, February 21, 2007 10:49 am, Eleanor Dodson wrote:
 There are many and various reasons for unpleasantly high R values,  and
 twinning is undoubtedly one of them.

 But you can only have twinning with apparently good diffraction if the
 cell dimensions allow it.

 It is always a possibility in trigonal, tetragonal and cubic cells.

 Certain other combinations of cell dimensions allow it ;- eg monoclinic
 with a ~= c   or beta ~= 90

 But you can usually detect it from the intensity statistics - see the
 plots from TRUNCATE for Moments and Cumulative intensities.

 Or the output of SFCHECK which suggests possible twinning baed on
 interpretation of  these tests.

 So it should be relatively easy to spot those cases where twinning is a
 likely cause of high Rfactor..

 Surprisingly some degree of twinning doesnt seem to degrade the map
 quality very much..

 Just an aside - it is really puzzling why from two sets of apparently
 similar data, one set gives Rfacros of 24+ while others  refine to R of
 18% ..
 Eleanor




 Kay Diederichs wrote:
 Sue Roberts wrote:
 Hello

 A partially philosophical, partially pragmatic question.

 I've noticed a trend, both on ccp4bb and locally, to jump to twinning
 as an explanation for data sets which do not refine well - that is
 data sets with R and Rfree stuck above whatever the person's
 pre-conceived idea of an acceptable R and Rfree are.   This usually
 leads to a mad chase through all possible space groups, twinning
 refinements, etc. and, in my experience, often results in a lot of
 time being spent for no significant improvements.

 Just out of curiosity, does  anyone have a feel for what fraction of
 stuck data sets are actually twinned? (I presume this will vary
 somewhat with the type of problem being worked on).

 And a  sorta-hypothetical question, given nice-looking crystals;
 images with no visible split spots, extra reflections, or streaks;
 good predictions; nice integration profiles; good scaling with
 reasonable systematic absences; a normal solvent content; and a
 plausible structure solution,  and R/Rf somewhat highish (lets say
 .25/.3  for 1.8 A data), how often would you expect the Stuck R/Rf to
 be caused by twinning (or would you not consider this a failed
 refinement).  (My bias is that such data sets are almost never
 twinned and one should look elsewhere for the problem, but perhaps
 others know better.)

 Sue
 Sue Roberts
 Biochemistry  Biopphysics
 University of Arizona

 [EMAIL PROTECTED]

 Sue,

 I seem to be in the other camp:  - nice-looking crystals; images
 with no visible split spots, extra reflections, or streaks; good
 predictions; nice integration profiles; good scaling with reasonable
 systematic absences; a normal solvent content; and a plausible structure
 solution,  and R/Rf somewhat highish (lets say .25/.3  for 1.8 A data)

 - all of this may happen with merohedrally twinned crystals. I believe
 it would be good to teach students to always devote some thought to
 the possibility of merohedral twinning in case of a trigonal/
 hexagonal/ tetragonal crystal, to avoid a rather common pitfall. I
 don't have the percentage at hand, but I believe I saw a paper by
 George Sheldrick giving a high percentage (like 20% or so) of
 merohedral twinned structures in the above crystal systems for
 small-molecule structures -  why should that percentage be different
 for protein crystals?

 It is of course true that twinning refinement is painful, and a lot of
 additional work! But man twinning is always enlightening reading.

 Kay

Re: [ccp4bb] REFMAC5 and Shannon factor

[Santarsiero, Bernard D.]

You didn't say anything about the weighting term between the F's and
geometrical parameters. That will substantially affect the R's, and the
default value of 0.3 in REFMAC isn't appropriate for all structures. In
CNS, it's adjusted to a more reasonable value during refinement.

Bernie Santarsiero


On Wed, February 14, 2007 10:36 am, Billy Poon wrote:
 Dear all,

 I have been using REFMAC5 to calculate the R values of several structures
 from
 the PDB and was confused by the behavior of the SHANnon_factor keyword.
 When I
 leave it at the default (set at 1.5 in the source code), I get one set of
 R
 values.  But when I manually set the value to 1.5 (SHANnon_factor 1.5) in
 the
 script, I get a different set of values.  Usually, it's off at the third
 or
 fourth decimal place, but for one case (PDB code 2OAU), I get a difference
 of
 about 0.0116 (0.28416 for default, 0.27258 for manual setting) for the R
 value
 and 0.0184 (0.31653 for default, 0.29818 for manual setting) in the Free
 R.

 I have tested this on the following architectures and they all behave the
 same
 way:

 Intel Xeon (32-bit Linux, REFMAC 5.2.0019, compiled with Intel compilers
 9.1)
 Intel Itanium2 (64-bit Linux, REFMAC 5.2.0019, compiled with Intel
 compilers
 9.1)
 AMD Opteron (64-bit Linux, REFMAC 5.2.0019, compiled with GNU compilers
 3.3.3)
 SGI MIPS (IRIX 6.5, REFMAC 5.2.0003, pre-compiled binaries)

 I am not doing any refining of the structures.  I just wanted to see what
 the R
 values are when calculated with REFMAC5 (the structures I was looking at
 were
 refined in CNS or X-PLOR) and was confused by the different results with
 the same setting.  Also, should changing the SHANnon_factor affect the R
 values much?  Thanks in advance for your help!

 -Billy

 And the script I use is pasted below:

 #
 # Refmac
 #
 refmac:
 refmac5 \
 HKLIN ./fo.mtz \
 XYZIN ./2oau.pdb \
 XYZOUT ./test.pdb \
  eor
 #
 # Do not add hydrogens
 #
 MAKE HYDR N
 MAKE LINK N
 MAKE CHECK NONE
 #
 # Input mtz labels
 #
 LABIN FP=FP SIGFP=SIGFP FREE=FREE
 #
 # Set resolution
 #
 REFI RESOlution 3.7 50.00
 #
 # Define NCS
 #
 NCSR NCHAI 7 CHAI A B C D E F G NSPANS 1 1 254 4
 #
 # Refine overall B factor
 #
 REFI BREF OVERall
 #
 # Set Free flag
 #
 FREE 0
 #
 # Number of refinement cycles
 #
 NCYC 0
 #
 # Monitoring level
 #
 MONI MEDI
 #
 # Change Shannon sampling (commented out if testing default behavior)
 #
 SHANNON_FACTOR 1.5
 #
 # end
 #
 end
 eor

Re: [ccp4bb] Multi-copies in one assymetric unit

[Santarsiero, Bernard D.]

Assuming that the MR solution is correct (I agree with Eleanor, that you
need to be certain about your selection of space group), then use your
complete model of chain A superimposed on chains B, C, and D. Then refine
and build the heck out of it. In a similar situation, it was relatively
easy for me to find two out of three molecules by MR. We were never really
convinced that there was a third (Matthews coefficient dicey, suggesting
either 2 or 3). After substantial refinement and building, the third chain
was apparent, and we rebuilt it completely. Looking at the average
B-factors of each chain, and the overall packing environment, chain A was
tight, chain B looser, and chain C will substantially fewer contacts.

Bernie Santarsiero


On Tue, February 13, 2007 10:07 am, Eleanor Dodson wrote:
 Sometimes this sort of disorder is due to an error , so the first thing
 is to check very carefully that the solution makes sense.

 Why are you so sure there are 6 copies in the asymmetric unit?

 In situations like this I first worry about SG.

 Is there a pseudo-translation vector? This can make it hard to decide on
 the SG..

 Is there an alternate spacegroup with fewer molecules in the asymm unit?

 What does the self rotation show?



 Yanming Zhang wrote:
 Dear All,

 Maybe this is a trivial question:

 My data should have 6 molecules in one assymetric unit.  MR could find
 out 4 molecules. After this, no matter how hard I have tried, no more
 molecules can be found. At this stage, I suppose that all other copies
 are dis-ordered. And go ahead to do refinement with 4 molecules (ABCD)
 available.
 The density for A is quite good. But for BCD are very dis-ordered.
 Many breaks in chains. I'd like to ask you:
 In a situation like this, should I:
 A, Use NCS for all copies?
 B Do not use NCS at all?
 C, Use NCS just for BCD? (even dis-ordered, but similar)?
 What is the trick to lower down Rfree as soon as possible, if you have
 experienced the same situation before?


 Thanks

 Yanming

Re: [ccp4bb] B in B-factor

[Santarsiero, Bernard D.]

page 79, chapter 5

On Wed, January 24, 2007 1:43 pm, Diana Tomchick wrote:
 Yes, it is! I somehow missed that when checking the IUCr web site.
 Virtually the whole publication is available as a series of
 downloadable PDF files, though for some strange reason the indices
 are listed but not available. Without the Name Index it would have
 taken a huge amount of time to locate the pertinent sections in a 700
 page book.

 Diana

 On Jan 24, 2007, at 1:23 PM, Miller, Mitchell D. wrote:

 Hi Diana,
 Is this the same publication you are referring to?
 http://www.iucr.org/iucr-top/publ/50YearsOfXrayDiffraction/index.html

 Regards,
 Mitch

 -Original Message-
 From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf
 Of Diana Tomchick
 Sent: Wednesday, January 24, 2007 11:09 AM
 To: CCP4BB@JISCMAIL.AC.UK
 Subject: Re: [ccp4bb] B in B-factor

 An interesting discussion of the development of the Debye-Waller
 correction can be found in the IUCr publication, Fifty Years of X-
 Ray Diffraction, edited by P.P. Ewald (unfortunately, it's no longer
 in print). It seems that Debye was the first scientist to tackle a
 problem which appeared to others hopelessly complicated at the time:
 the influence of the temperature motion of the atoms on the
 diffraction of X-rays (p. 79 of aforementioned publication).
 Interestingly, his original formulation of the correction included
 the constant M, which can be expressed by means of the elastic
 properties of the crystal or their combination, the 'Debye
 Temperature', which occurs in the theory of the specific heat (same
 reference). The text goes on to mention that Ivar Waller
 demonstrated in 1923 that Debye's expression for M was wrong by a
 factor of two. No mention is given in Fifty Years of a B factor
 but it sounds as though Waller decided to use a different
 nomenclature in his paper from 1927 to avoid confusion.

 Too bad that Fifty Years is no longer in print, though used copies
 can be purchased from various on-line booksellers. It provides a very
 enlightening and easily readable history of the early days in the
 field, and the interchange of scientists and ideas across the
 Atlantic Ocean.

 Diana


 On Jan 24, 2007, at 11:14 AM, Santarsiero, Bernard D. wrote:

 It looks like the earliest reference to the Debye-Waller factor is
 from
 Debye's paper:

 Uber den Einfluss der Warmebewegung uf die Interferenzerscheinungen
 beiu
 Rontgenstrahlen, Verhandl. deut. phyik. Ges., 15, 678-689 (1913),

 and the succeeding paper Verhandl. deut. phyik, Ges., 15, 738-752
 (1913),

 and Waller:

 Die Einwirkung der Warmbewegung der Kristllatome auf Intersitat,
 Lage, and
 Scharfe der Rontgenspektrallinein, Ann. Physik, 83, 153-183 (1927).

 Interestingly, beta was used for the anisotropic Gaussian form of the
 isotropic Gaussian B-factor.

 Bernie



 On Wed, January 24, 2007 9:05 am, Roberto Steiner wrote:
 Hi Rajesh,

 I also wondered about that.
 Not having been able to find a good explanation in the literature
 I thought of it as reducing factor of the *B*ragg peaks. M???

 Then I stopped thinking about it (I now try to call them ADPs)


 Roberto

 On 24 Jan 2007, at 12:42, Rajesh Kumar Singh wrote:

 May be too trivial, I was just wondering
 what B stands for in the term B-factor.

 Thanks

 Rajesh

 --
 Rajesh Kumar Singh
 Institut fur Biochemie
 Universitat Greifswald
 Felix-Hausdorff-Str. 4
 D-17489 Greifswald
 Germany

 E.Mail: [EMAIL PROTECTED]
 Phone: +49-3834- 86 4392

 ---
 Dr. Roberto Steiner
 Randall Division of Cell and Molecular Biophysics
 New Hunt's House
 King's College London
 Guy's Campus
 London, SE1 1UL
 Phone +44 (0)20-7848-8216
 Fax   +44 (0)20-7848-6435
 e-mail [EMAIL PROTECTED]


 * * * * * * * * * * * * * * * * * * * * * * * * * * * *
 Diana R. Tomchick
 Associate Professor
 University of Texas Southwestern Medical Center
 Department of Biochemistry
 5323 Harry Hines Blvd.
 Rm. ND10.214B
 Dallas, TX 75390-8816, U.S.A.
 Email: [EMAIL PROTECTED]
 214-645-6383 (phone)
 214-645-6353 (fax)

 * * * * * * * * * * * * * * * * * * * * * * * * * * * *
 Diana R. Tomchick
 Associate Professor
 University of Texas Southwestern Medical Center
 Department of Biochemistry
 5323 Harry Hines Blvd.
 Rm. ND10.214B
 Dallas, TX 75390-8816, U.S.A.
 Email: [EMAIL PROTECTED]
 214-645-6383 (phone)
 214-645-6353 (fax)