[phenixbb] problem with refinement in AutoBuild

Michael Thompson miket at chem.ucla.edu
Thu Aug 15 13:55:49 PDT 2013

Hello Everyone,

I am having a problem running autobuild, which appears to be happening during the refinement step. I start autobuild using :

phenix.autobuild data=data.mtz model=model.mtz seq_file=sequence.dat rebuild_from_fragments=True i_ran_seed=124889 find_ncs=False refine_with_ncs=False nproc=4

The runs are failing during the first cycle with the following error message: 

AutoBuild_build_cycle  AutoBuild  Run 2 Thu Aug 15 13:37:34 2013

Build cycle 1 of 16   method:rebuild
Initial rebuild cycle: just combining previous models or copying 
over initial PDB file

Using NO_LIG_NO_WAT_CloMP_AutoBuild_input_ed.pdb for rebuild, copied to unrefined.pdb_1
Refining model:  unrefined.pdb_1
Refinement strategy:  ['individual_sites', 'individual_adp', 'occupancies']
No refinement carried out (may need free R flags)
Error message from phenix.refine:  
Refinement failed...
Failed to carry out AutoBuild_build_cycle:

Initial phenix refinement of AutoBuild_run_2_/TEMP0/NO_LIG_NO_WAT_CloMP_AutoBuild_input_ed.pdb with aniso_refinement_data_PHX.mtz failed...quitting...
/home/miket/projects/CloMP/Oxidized/APSoct2012/CloMP-TMAOcryo_B2_3/MRwithChainsAB/RefineCChalf/AutobuildPrep/AutoBuild_run_2_/LAST.LOG may have additional information...


I also checked the LAST.LOG file that is referenced in the error message. That appears to be a phenix.refine logfile, however it just cuts off after the "refinement start" header, with no information about the cause of the failure (no error message).

My input MTZ file does indeed contain R-free flags. I have used this MTZ for refinement with phenix.refine (outside of autobuild) with no problems. Furthermore, earlier in the autobuild logfile, I see: 

Copying  refinement_PHX.mtz  to  exptl_fobs_phases_freeR_flags.mtz
Columns used:  FP=FP SIGFP=SIGFP FreeR_flag=FreeRflag

Which seems to me like it is recognizing my R-free flags column in the input MTZ. 

Does anyone have any idea what might be going wrong here??



Michael C. Thompson

Graduate Student

Biochemistry & Molecular Biology Division

Department of Chemistry & Biochemistry

University of California, Los Angeles

miket at chem.ucla.edu

More information about the phenixbb mailing list