2022.3

• Added mmtbx.process_predicted_model command
• Updates and bug fixes to cctbx.HKLviewer
• Added methods for improved handling of heavy hydrogens in model

2022.2

• Added --quiet flag to CCTBXParser to suppress output
• Updates to restraints based on quantum mechanics
• Updates and bug fixes to cctbx.HKLviewer

2022.1

• Added option to use legacy bulk solvent mask
• Added option to any_reflection_file_reader to control averaging of anomalous data columns in MTZ files
• Fix SHELXF formatting where integer values may be interpreted as decimal

2021.12

• Initial support for Python 3.10
• Added support for outputting multi-model mmCIF files
• Adjusted mask gridding for bulk solvent

2021.11

• Added right-handed nucleic acids for DNA
• Improved handling of different unit cells in MTZ file
• Avoid division by zero when rotating 0°
• Added option to ignore secondary strucure annotations when reading
  models through the DataManager
• Initial support for restraints based on quantum mechanics
• Improved consistency of binning by d_star_sq

2021.10

• Initial migration of MolProbity functionality (probe and reduce) to mmtbx
• Initial tool for likelihood-based map preparation for docking
• Improvements to ADP refinement for real-space

2021.9

• Improved structure factor calculation at ultra-low resolution
• Improved processing of prediced models
• Added diffBragg to simtbx for modeling pixels in stills to improve structure factors
• Added suitename to mmtbx for classifying RNA

2021.8

• Added tools for processing predicted models based on the error estimate
• Updated list of modified amino and nucleic acids
• Better handling of sequence files with empty sequences

2021.7

• Fix pickling error with anomalous_probability_plot
• Fix bug in reading data CIF file with paired data and sigma arrays of
  different sizes
• Added functions for retrieving H-bond types and Van der Waals radii to
  the model manager class
• Sequence validation will only use protein or nucleic acid residues for
  alignment

2021.6

• More improvements to bulk solvent masking for multiple regions
• Updates to ensemble refinement
• Enable conversion of some numpy types into flex types instead of
  requiring that the types match (e.g. int to float is now supported)