[phenixbb] phenix autobuild question

aberndt at mrc-lmb.cam.ac.uk aberndt at mrc-lmb.cam.ac.uk
Mon Sep 15 13:45:24 PDT 2008


Hi Tom,

thanks for your answer. I just checked the phenix version I used and it is
version-1.3-final. Please find attached the .log file you asked for. I
just realised that resolve apparently had a problem with the .mtz label
assignment. I never had this problem before. Now that I mention that I
have to confess that this is not entirely true. All the phenix tutorials I
have looked through so far state that you can use ccp4s mtz the way they
are. Yet, this is not entirely true. Say I want to have an analysis from
phenix.xtriage by typing 'phenix.xtriage latest.mtz'. So phenix will give
me the error message:
Multiple equally suitable arrays of observed xray data found.

Possible choices:
  ab_xds_pointless_scala5.mtz:IMEAN_XDSdataset,SIGIMEAN_XDSdataset
  ab_xds_pointless_scala5.mtz:I_XDSdataset(+),SIGI_XDSdataset(+),I_XDSdataset(-),SIGI_XDSdataset(-),merged

Please use scaling.input.xray_data.obs_labels
to specify an unambiguous substring of the target label.

Sorry: Multiple equally suitable arrays of observed xray data found.

Is there a way a way to avoid going back to ccp and sorting the labels
accordingly. What would be the 'scaling.input.xray_data.obs_labels'
command in the xtriage command line?

Anyway, Tom, many thanks for your answer in advance! I reaaly like working
with phenix. I just need to learn loads more ...

Best,
Alex



> Hi Alex,
>
> I'm sorry for both the failure and the lack of a clear message here!  I
> will try to fix both.
>
> Is this running phenix-1.3-final?  (If not, could you possibly run with
> that version so we are on the same version.)
>
> Can you possibly send me the end of the output in
>
> /someplace/home/someuser/
> work/someproject/phenix2/AutoBuild_run_1_/TEMP0/AutoBuild_run_1_/AutoBuild_run_1_1.log
>
> which I hope will have an actual error message to look at?
>
> The reason for the subdirectories is this: phenix.autobuild runs several
> jobs in parallel (if you have set nproc ) or one after the other (if you
> have not set nproc).  These subdirectories contain those jobs...which are
> then combined together to give the final results in your overall
> AutoBuild_run_1_/ directory.
>
> I don't know the answer to the phenix.refine question...perhaps Pawel or
> Ralf can answer that one.
>
> All the best,
> Tom T
>
>> Dear phenix community,
>>
>> I started an AutoBuild run using the phenix GUI (ver 1.3 rc4) on a
>> linux cluster. It worked okay and all the refinement looked pretty
>> decent. However, after quite a while I obtained the following error
>> message (see below). Would anyone please tell me what to do to prevent
>> this error and get phenix get its job finished?
>>
>> Also, I don't understand why phenix AutoBuild creates 5 (or whatever
>> number) AutoBuild_run_x in the AutoBuild-run_1/TEMP0 directory and not
>> two dirs up in the tree. In other words in my /someplace/home/someuser/
>> work/someproject/phenix2/AutoBuild_run_1_/TEMP0 directory are  more
>> AutoBuild_run_x_ directories (with x=1 to 5). It is a bit confusing to
>> me.
>>
>> A final question: I realised that the phenix.refine drastically
>> increases the number of outliers. I know that thare is a weighting
>> term someplace ... but what was it again?
>>
>> Many thanks in advance,
>> Alex
>>
>>
>> warnings
>> Failed to carry out AutoBuild_multiple_models:
>> Sorry, subprocess failed...message is:
>> ********************************************************************************
>> phenix.autobuild   \
>>
>> write_run_directory_to_file=/someplace/home/someuser/work/someproject/
>> phenix2/AutoBuild_run_1_/TEMP0/INFO_FILE_1
>>
>> Reading effective parameters from /someplace/home/someuser/work/
>> someproject/phenix2/AutoBuild_run_1_/TEMP0/PARAMS_1.eff
>>
>> Sending output to  AutoBuild_run_1_/AutoBuild_run_1_1.log
>>
>> ********************************************************************************
>>
>> Failed to carry out AutoBuild_build_cycle:
>>
>> failure
>>
>> ********************************************************************************
>>
>> ********************************************************************************
>>
>> work_pdb_file  "AutoBuild_run_1_/edited_pdb.pdb"
>> working_directory  "/someplace/home/someuser/work/someproject/phenix2"
>> worst_percent_res_rebuild  2.0
>>
>> ---------------------------------------------------
>> Alex Berndt
>> MRC-Laboratory of Molecular Biology
>> Hills Road
>> Cambridge, CB2 0QH
>> U.K.
>>
>> phone: +44 (0)1223 402113
>> ---------------------------------------------------
>>
>> _______________________________________________
>> phenixbb mailing list
>> phenixbb at phenix-online.org
>> http://www.phenix-online.org/mailman/listinfo/phenixbb
>>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: AutoBuild_run_1_1.log
Type: application/octet-stream
Size: 11674 bytes
Desc: not available
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20080915/609e1cf3/attachment-0001.obj>


More information about the phenixbb mailing list