[phenixbb] Molecular replacement pipeline MRage
gb360 at cam.ac.uk
Mon Aug 5 04:25:43 PDT 2013
your comments about more transparency are well received!
To reassure you, the logfile is for users and not for the program (in fact,
the program does not rely on this file being available, and it is only
saved for archival purposes). However, I start realising that we need to
add some recovery mechanism to resume after a crash. Although all results
are printed immediately after received from the underlying calculation, it
is not in a machine readable format, and right now manual after a crash
recovery needs to be done manually. In addition, in the near future, there
will be some improvements in logfile reporting, with the program displaying
the current percentage of discovery regularly, so that users can have a
vague idea of how long this is going to take. I will be happy to add a list
of best peaks founds so far to this, if it is potentially useful.
Regarding your latter comment, could you clarify what happened? The way I
understand you is that you ran a separate Phaser task from the Phenix GUI,
and it crashed, and took the GUI (and everything else running in the GUI,
i.e. MRage in your case). Is this correct?
On Aug 5 2013, Morten Grøftehauge wrote:
>So I've been trying out MRage. I used the GUI since I didn't care too much
>and just wanted it started.
>I noticed two things. It doesn't show you the best results so far. It
>doesn't even save the best results so far. Everything is hidden in a giant
>log file in the .phenix/project_data folder (.phenix is hidden so I'm
>guessing that means you don't want people snooping around in there?).
>I can grep the TFZ= lines with grep 'TFZ=' mr_pipeline.log >
>grep-TFZ.txtand sort them with sort
>-r -k 13 grep-TFZ.txt but I think this is a bit excessive for the average
>Oh, and now I tried to re-run the highest scoring model in Phaser and it
>crashed and took MRage with it. Are you sure MRage is out of beta? Trying
>to be facetious, not mean.
More information about the phenixbb