Hi Tim, I think we ignore arrays with *all* zeros (as Nat pointed out), not just select Miller index entries that have data value zero. That is if you have 1000 of Fobs and all 1000 of them are equal to zero such an array will be ignored. Perhaps this also depends on the specific program. All the best, Pavel On 5/12/14, 8:54 AM, Tim Gruene wrote:

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1

Dear Nat,

I don't know the formula for HL by heart and I don't see the Miller-Indices, but couldn't centric reflections have all four HLs = 0 if the two-fold ambiguity is resolved and those reflections have phi=0? I thought phi=0 corresponds to A=B=C=D=0.

Best, Tim

On 05/12/2014 05:35 PM, Nathaniel Echols wrote:

...

Here is the problem:

phenix.mtz.dump resolve_1_190_phenix.mtz ... HLA xxxxx 100.00% 0.00 0.00 A: phase probability coefficients (Hendrickson/Lattman) HLB xxxxx 100.00% 0.00 0.00 A: phase probability coefficients (Hendrickson/Lattman) HLC xxxxx 100.00% 0.00 0.00 A: phase probability coefficients (Hendrickson/Lattman) HLD xxxxx 100.00% 0.00 0.00 A: phase probability coefficients (Hendrickson/Lattman)

We ignore arrays with all zeros, since they're obviously not correct - however, the feedback given to the user is deficient. I am open to suggestions about how we can improve this, but I am reluctant to assault users with pop-ups every time bogus information is detected, especially since this will increase the latency of the GUI.

The file from AutoSol you should actually use is probably named overall_best_refine_data.mtz, which contains the original, purely experimental phases (along with data and R-free flags, of course). We don't recommend using the density-modified phases as restraints. (I am not sure why Resolve would output all zeros; that is a question for Tom. I did however notice the same thing in a file generated by AutoSol for a tutorial dataset.)

-Nat

On Mon, May 12, 2014 at 8:21 AM, Jonathan Grimes wrote:

...

hi nat

i created this mtz file.....by taking a resolve.mtz and putting it back thru phenix.

from the GUI and from the command line phenix doesnt like my HLs coeffs.

thanks jon

Dr. Jonathan M. Grimes, NDM Senior Reseach Fellow Diamond Research Fellow University Research Lecturer Diamond Light Source Division of Structural Biology Diamond House Wellcome Trust Centre for Human Genetics Harwell Science Campus University of Oxford Didcot Roosevelt Drive, OX11 0DE Oxford OX3 7BN, UK

Email: [email protected], Web: www.strubi.ox.ac.uk Tel: (+44) - 1865 - 287561, FAX: (+44) - 1865 - 287547

---- Original message ----

...

Date: Mon, 12 May 2014 07:44:14 -0700 From: Nathaniel Echols Subject: Re: [phenixbb] HLs not recognised To: Jonathan Grimes Cc: PHENIX user mailing list

This is a little strange - the GUI should not be allowing you to select an array that won't work for refinement. Could you please send me (off-list!) both MTZ files that you tried using? (I don't think I need data or anything else for this.) thanks, Nat

On Mon, May 12, 2014 at 5:49 AM, Jonathan Grimes wrote:

i have now used phenix to read in the mtz file, read all the arrays and make a new mtz file.

phenix recognised the HLs as Hend/latt coeffs.

i used the eff file to run phenix.refine at the command line and got this error

Not a suitable array of experimental phases: refinement.input.experimental_phases.labels=HLA,HLB,HLC,HL

The phases were orginally derived by Phenix using the auto solve GUI. Jon

Dr. Jonathan M. Grimes, NDM Senior Reseach Fellow Diamond Research Fellow University Research Lecturer Diamond Light Source Division of Structural Biology Diamond House Wellcome Trust Centre for Human Genetics Harwell Science Campus University of Oxford Didcot Roosevelt Drive, OX11 0DE Oxford OX3 7BN, UK

Email: [email protected], Web: www.strubi.ox.ac.uk Tel: (+44) - 1865 - 287561, FAX: (+44) - 1865 - 287547

...

Date: Mon, 12 May 2014 11:32:39 +0100 From: [email protected] (on behalf of Jonathan Grimes ) Subject: [phenixbb] HLs not recognised To: [email protected]

I am trying to refine a structure at low resolution and would like to use phase restraints. using the GUI even

---- Original message ---- though HLs are present

...

in the MTZ file, and are type ?A??checked with CAD?..i am not given the option of selecting them.

not clear to me what I?m doing wrong.

hints most wellcome thanks jon

Dr. Jonathan M. Grimes, NDM Senior Reseach Fellow University Research Lecturer DIAMOND Research Fellow

Division of Structural Biology Wellcome Trust Centre for Human Genetics University of Oxford Roosevelt Drive, Oxford OX3 7BN, UK

Email: [email protected], Web: www.strubi.ox.ac.uk Tel: (+44) - 1865 - 287561, FAX: (+44) - 1865 - 287547 _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

- -- - -- Dr Tim Gruene Institut fuer anorganische Chemie Tammannstr. 4 D-37077 Goettingen

GPG Key ID = A46BEE1A

-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Icedove - http://www.enigmail.net/

iD8DBQFTcO60UxlJ7aRr7hoRAtWhAJsE0yoPEDA3g12VIS5KRu7pdDdCbACgkK5N 2bvVsUVRhTBmXgXzyA7FADQ= =n9Sj -----END PGP SIGNATURE----- _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

Hi, Just to be a bit more precise about that — the part of the phase information that comes from the two hands of the Friedel pair is missing for centric reflections in SAD phasing, but the part of the phase information that comes from knowing even a small part of the structure (i.e. the real scattering of the anomalous substructure) is still present. So, in general, the HL coefficients should not all be zero for the centric reflections in Phaser SAD phasing output, but they will generally be smaller (and the figures of merit will also be smaller) than for the acentric reflections. The same goes for singletons (our term for the acentric reflections where only F+ or F- but not both have been measured). Randy On 12 May 2014, at 17:25, Nathaniel Echols wrote:

On Mon, May 12, 2014 at 8:54 AM, Tim Gruene wrote: I don't know the formula for HL by heart and I don't see the Miller-Indices, but couldn't centric reflections have all four HLs = 0 if the two-fold ambiguity is resolved and those reflections have phi=0? I thought phi=0 corresponds to A=B=C=D=0.

Based on my reading of the ccp4 wiki, this appears to be incorrect:

http://ccp4wiki.org/~ccp4wiki/wiki/index.php?title=Hendrickson_Lattman_coeff...

C=D=0 would be consistent with the phase ambiguity being resolved, but if A=B=0 that would imply that the figure of merit is also zero. However, it turns out that the tutorial output I was looking at does have all centric HL coefficients equal to zero - but my understanding (possibly incorrect) was that the centric reflections can't be phased using SAD (which explains why the figure of merit for centrics is always really poor in the Phaser output). In any case, if the MTZ file contains only centric reflections that by itself is worrying, although I found some lovely examples of this in the PDB yesterday.

-Nat _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

------ Randy J. Read Department of Haematology, University of Cambridge Cambridge Institute for Medical Research Tel: + 44 1223 336500 Wellcome Trust/MRC Building Fax: + 44 1223 336827 Hills Road E-mail: [email protected] Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk