search

xchem / XChemExplorer

6 stars 8 forks source link

Edstats failure (TBRU data) leading to no ligand validation plot #118

Closed nelse003 closed 4 years ago

nelse003 commented 6 years ago

Explore reason, and turn into actionable item.

File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/runpy.py", line 162, in _run_module_as_main "main", fname, loader, pkg_name) File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/runpy.py", line 72, in _run_code exec code in run_globals File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/site-packages/giant/jiffies/score_model.py", line 497, in description = DESCRIPTION) File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/site-packages/giant/jiffies/init.py", line 16, in init out = run(params=working_phil.extract()) File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/site-packages/giant/jiffies/score_model.py", line 456, in run verbose = params.settings.verbose File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/site-packages/giant/jiffies/score_model.py", line 211, in score_model mtz1_edstats_scores = Edstats(mtz_file=mtz1, pdb_file=pdb1) File "/dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/lib/python2.7/site-packages/giant/xray/edstats.py", line 18, in init raise Exception('EDSTATS ERROR\n=========>\n{!s}\n=========>'.format(self._command.error)) Exception: EDSTATS ERROR =========>

WARNING: Experiment type is not defined, assuming 'X-RAY DIFFRACTION' edstats: ERROR: Either PDBOUT or CIFOUT must be specified!

ERROR in edstats, at /dls/science/groups/i04-1/software/pandda-update/ccp4-linux64-2016-10-22-0115/bin/edstats.pl line 361.

=========>

nelse003 commented 6 years ago

This relates to the new version of EDSTATS supplied specifically in the DLS CCP4: See #92. The required outputs are not being provided by

Options:

  1. Nick to update optional arguments to giant.xray.edstats score_with_edstats_to_list(mtz_file, pdb_file) to allow parsing both new and old version of edstats.

  2. Elliot to do above, test and push to pandda hotfix.

  3. Revert EDSTATS.

nelse003 commented 6 years ago

EDSTATS was updated in the latest CCP4 release. Shouldn’t that fix this problem?

Nick

That's what's causing this problem, the updated version of edstats requires a larger number of command line arguments.

i.e. current giant.score_model works on a ccp4 downloaded recently, but without the new edstats perl script. Does not work on ccp4 sourced by xce that has updated edstats release.

1) I don’t know why the new version should suddenly demand a PDBOUT — maybe drop Ian a quick email to check that’s intended functionality or bug. There is functionality to output values into the occupancy column of the PDB file, so he may have accidentally deleted an if statement or something, meaning that it’s always required. 2) Question: Is the edstats released in CCP4 the other day the same as the version Ian sent you with the fix? (i.e. does it have the chain-assignment fix?). 3) I don’t think the two things — 1) and 2) — should be related: I thought that Ian’s updated version automatically dealt with the funny chain-assignment thing without extra command line stuff… 4) Elliot: If required, maybe you should have a go at changing the code in giant.xray… I suspect you just need to direct "pdbout" to a dummy temp file and delete it after edstats is run. Shouldn’t require a flag to switch between versions.