Skip to content
Snippets Groups Projects
  • Studer Gabriel's avatar
    8f0d268d
    Improve compression in OMF · 8f0d268d
    Studer Gabriel authored
    Adds several features that can all be enabled separately:
    - DEFAULT_PEPLIB: OMF stores a ResidueDefinition object for each observed
                      unique compound. No need to dump residue definitions for
                      standard amino acids.
    - LOSSY: Reduce coordinate accuracy to 0.1
    - AVG_BFACTORS: Optimization specific for models, where all atoms of a residue
                    have the same bfactor => store one bfactor per residue instead
                    of one bfactor per atom
    - ROUND_BFACTORS: Round bfactors, i.e. 42.42 => 42.0
    - SKIP_SS: Don't dump secondary structure, loaded objects assign COIL to
               each residue
    - INFER_PEP_BONDS: Intra-residue bonds are stored in ResidueDefinition objects.
                       Inter-residue bonds need to be stored separately. No need
                       to do that for peptide bonds as this can be inferred on the
                       fly when loading
    8f0d268d
    History
    Improve compression in OMF
    Studer Gabriel authored
    Adds several features that can all be enabled separately:
    - DEFAULT_PEPLIB: OMF stores a ResidueDefinition object for each observed
                      unique compound. No need to dump residue definitions for
                      standard amino acids.
    - LOSSY: Reduce coordinate accuracy to 0.1
    - AVG_BFACTORS: Optimization specific for models, where all atoms of a residue
                    have the same bfactor => store one bfactor per residue instead
                    of one bfactor per atom
    - ROUND_BFACTORS: Round bfactors, i.e. 42.42 => 42.0
    - SKIP_SS: Don't dump secondary structure, loaded objects assign COIL to
               each residue
    - INFER_PEP_BONDS: Intra-residue bonds are stored in ResidueDefinition objects.
                       Inter-residue bonds need to be stored separately. No need
                       to do that for peptide bonds as this can be inferred on the
                       fly when loading