1.19.1 Changes

• phenix.real_space_refine
• Fix bug in restoring previous jobs
• Clarify NCS options

• phenix.real_space_refine
• Improved rotamer fitting (use multiprocessing, in case of NCS constraints work on one copy only and propagate changes to all related copies, various performace improvements and bug fixes)
• Improved map/restraints weight calculation
• Morphing can now use multiprocessing (nproc)
• phenix.map_to_model and phenix.trace_and_build
• Improved high-resolution model-building including detection of insertions and deletions
• New methods
• phenix.local_resolution: calculates a local resolution map
• phenix.local_aniso_sharpen: optimizes a map taking into account local resolution-dependence and anisotropy of the map and its errors
• New scripting tools
• High-level scriptable Python tools are now available for map and model analyses and manipulation as well as for model-building
• Restraints
• Adjusting the "positions" of atom names is (pseudo-)symmetric amino acid side chains is now the default
• Improved restraints for ARG allows more flexibility of the CD atom
• Bug fixes
• Chains with modified amino or nucleic acids can be aligned
• Fixed handling of modified amino/nucleic acids in sequence alignment

1.18.2 Changes

• Fix weighting for iron sulfur clusters

1.18.1 Changes

• General bug fixes

• Amber
• Automatic creation of Amber files in phenix.refine GUI
• Added AmberPrep GUI
• Restraints
• Engh & Huber restraints (2001) for cis-peptides now implemented
• Restraints added for FeS
• Metal coordination library is default to Zn+2 and FeS clusters
• Rama-Z: New global Ramachandran score for identifying protein structures with unlikely stereochemistry in CLI (phenix.rama_z, mmtbx.rama_z) and GUI (validation reports).
• Density modification for cryo-EM
• Includes model-based density modification with automatic model generation
• Optimized defaults and additional documentation
• Real-space refinement
• Hydrogen atoms no longer included in map target function, which improves fit.
• Add NQH flip option (enabled by default)
• Other
• New tool for ordered solvent picking in cryo-EM maps (phenix.douse)
New map and model superposition tool (phenix.match_maps)
• New FindProgram toll to find any Phenix program with a text search
• Project details now has a button for opening the README file for tutorials

1.17.1 Changes

• Fix bug on Windows where Xtriage results would fail to display
• Other bug fixes

• Improved handling of SHELX data in phenix.reflection_file_converter
• eLBOW can output files for Amber
• eLBOW supports QM package Orca
• dials.image_viewer is used for viewing diffraction images
• Updated map smoothing
• Fix inconsistency in clashscore values in phenix.validation_cryoem when hydrogens are in the model

1.16 Changes

• New GUI for mmtbx.prepare_pdb_deposition to make mmCIF files for deposition into the PDB
• New tool (CLI and GUI) for getting a validation report from the PDB (phenix.get_pdb_validation_report)
• Added sequence checking to Comprehensive Validation for Cryo-EM
• Amber functionality is available by default

1.15.2 Changes

• Fix phenix.rosetta_refine issue
• Fix PDB/CIF conversion in GUI

1.15.1 Changes

• Improved phenix.map_to_model behavior

New and improved tools

• New algorithm for phenix.map_to_model is faster and builds longer chains
• phenix.trace_and_build can build protein into maps at resolutions as low as 4.5 A
• phenix.fix_insertions_deletions can build models in places where the map is poor
• phenix.refine_ca_model for optimizing C-alpha only models
• phenix.comparama for generating Kleywegt-like plots that show how residues moved in the Rakmachandran plots before and after refinement
• eLBOW can find unique instances of a ligand from the PDB and optionally create Polder OMIT maps
• Updated structural library for phenix.structure_search
• Updated ligand library for phenix.ligand_identification
• Improved reporting of cis-peptides for residues with altloc atoms

• Updated mmCIF support (stuct_conn loop, ligand restraints, sequence)
• phenix.refine, phenix.real_space_refine, and phenix.pdbtools will also output models in mmCIF format by default
• Dependencies are now based on conda packages, which will be more compatible with new operating systems and will improve consistency across all platforms (macOS, linux, Windows)
• Bug fixes