(Bug/feature?) Protein chain too short!

Issue #21 open
Maciej Ciemny
created an issue

Command

cabsDock -r 1Z9L:A -p EDEFYDALS:CCCCEECCC --work-dir 1Z9L --reference-pdb 1Z9O:ACG --contact-maps  --image-file-format png

For unknown reasons raises exception „Exception: Protein chain too short!". The 1z9l chain A looks OK when viewed in PyMol.

CABS-docking job 1Z9L:A
 Building complex...
['EDEFYDALS:CCCCEECCC', 'random', 'random']
 ... done.
Traceback (most recent call last):
  File "/home/users/mciem/.local/bin/cabsDock", line 11, in <module>
    load_entry_point('cabsDock==1.0', 'console_scripts', 'cabsDock')()
  File "build/bdist.linux-x86_64/egg/cabsDock/__main__.py", line 65, in run_job
  File "build/bdist.linux-x86_64/egg/cabsDock/job.py", line 197, in cabsdock
  File "build/bdist.linux-x86_64/egg/cabsDock/job.py", line 242, in setup_cabs_run
  File "build/bdist.linux-x86_64/egg/cabsDock/cabs.py", line 114, in __init__
  File "build/bdist.linux-x86_64/egg/cabsDock/cabs.py", line 192, in load_structure
  File "build/bdist.linux-x86_64/egg/cabsDock/cabs.py", line 58, in cast
Exception: Protein chain too short!

Comments (11)

  1. Mateusz Kurciński

    fixed #19 user must now provide reference-pdb as <code/file>:<rec chains>:<pep chains> fixed #36 fixed #21 non-standard substituion moved to PDBlib BEFORE applying selection fixed #38

    ca2all module: init now accepts argument work_dir where temp files are generated [default is '.'] option modeller-iterations now does what it says, default is 3 * family.mat is now <temp_prefix>.mat

    → <<cset 716807070ad0>>

  2. Mateusz Kurciński

    fixed #19 user must now provide reference-pdb as <code/file>:<rec chains>:<pep chains> fixed #36 fixed #21 non-standard substituion moved to PDBlib BEFORE applying selection fixed #38

    ca2all module: init now accepts argument work_dir where temp files are generated [default is '.'] option modeller-iterations now does what it says, default is 3 * family.mat is now <temp_prefix>.mat

    → <<cset 709aef928ce2>>

  3. Maciej Ciemny reporter
    • changed status to open

    This issue is still an issue:

    Running:

    CABSflex -i 3uvv --work-dir 3uvv -v 4 --remote
    

    (other example that results in the same error is 4y2h)

    Results in:

    [INFO]       Logger:   Verbosity set to: 4 - [DEBUG]                                               (00:00:08)
    [WARNING]    JOB:      /STORAGE/DATA/mciem/flexmultichain_benchrun_Fri_Nov_17_10:46:30_2017/run/3u
                           vv already exists. Output data will be overwritten.                         (00:00:09)
    [DEBUG]      Protein:  Preparing the complex                                                       (00:00:09)
    [DEBUG]      PDB:      Creating Pdb object from 3uvv                                               (00:00:09)
    [DEBUG]      PDB:      Downloading 3uvv                                                            (00:00:09)
    [DEBUG]      PDB:      Processing 3uvv                                                             (00:00:14)
    [DEBUG]      PDB:      Removing alternative locations from 3uvv                                    (00:00:14)
    [DEBUG]      PDB:      Removing water molecules from 3uvv                                          (00:00:14)
    [DEBUG]      PDB:      Scanning 3uvv for non-standard amino acids                                  (00:00:15)
    [WARNING]    PDB:      Unknown residue T3 at 501:A in 3uvv                                         (00:00:15)
    [WARNING]    PDB:      Unknown residue 9CR at 501:B in 3uvv                                        (00:00:15)
    [DEBUG]      PDB:      Removing heteroatoms from 3uvv                                              (00:00:15)
    [DEBUG]      PDB:      Selecting [name CA] from 3uvv                                               (00:00:15)
    [INFO]       Protein:  Loading 3uvv as input protein                                               (00:00:15)
    [DEBUG]      PDB:      Running DSSP                                                                (00:00:15)
    [DEBUG]      PDB:      DSSP successful                                                             (00:00:15)
    [OUT FILES]  Logger:   Saving DSSP output to /STORAGE/DATA/mciem/flexmultichain_benchrun_Fri_Nov_1
                           7_10:46:30_2017/run/3uvv/output_data/DSSP_output_3uvv.txt                   (00:00:15)
    [WARNING]    Protein:  Could not read weights file: gauss                                          (00:00:16)
    [WARNING]    Protein:  Using default weights(1.0) for all atoms.                                   (00:00:16)
    [DEBUG]      Protein:  Complex successfully created                                                (00:00:17)
    [INFO]       CABS:     Setting up CABS simulation.                                                 (00:00:17)
    [DEBUG]      CABS:     Loading structures...                                                       (00:00:17)
    [CRITICAL]   CABSflex: Error occured: Protein chain too short!                                     (00:00:18)
    Traceback (most recent call last):
      File "/home/users/mciem/.local/lib/python2.7/site-packages/CABS/__main__.py", line 84, in run_flex
        job.run()
      File "/home/users/mciem/.local/lib/python2.7/site-packages/CABS/job.py", line 166, in run
        self.setup_cabs_run()
      File "/home/users/mciem/.local/lib/python2.7/site-packages/CABS/job.py", line 327, in setup_cabs_run
        mc_steps=self.mc_steps
      File "/home/users/mciem/.local/lib/python2.7/site-packages/CABS/cabs.py", line 122, in __init__
        fchains, seq, ids = CabsRun.load_structure(protein_complex)
      File "/home/users/mciem/.local/lib/python2.7/site-packages/CABS/cabs.py", line 199, in load_structure
        vectors = CabsRun.LATTICE.cast(chain)
      File "/home/users/mciem/.local/lib/python2.7/site-packages/CABS/cabs.py", line 64, in cast
        raise Exception('Protein chain too short!')
    Exception: Protein chain too short!
    
  4. Mateusz Kurciński

    https://bitbucket.org/lcbio/cabsdock/commits/b230b8a059c391453db9edc71cce5d431bdaec11

    Zmieniłem funkcję do znajdowania dziur w łańcuchach. Teraz jak powstanie łańcuch krótszy niż 4 reszty to jest usuwany. Potencjał do 'fuckupu' w innych miejscach jest spory, bo nie jestem w stanie szybko prześledzić, czy takie podejście ma jakieś konsekwencje do tego co się dzieje dalej. @Maciej Ciemny proszę Cię o puszczenie tej wersji na jakimś testowym zestawie (te zmiany są na razie tylko w develop), w release i master jest ostatnia wersja 0.9.12

  5. Log in to comment