Skip to content

Latest commit

 

History

History
235 lines (146 loc) · 12.2 KB

NEWS.md

File metadata and controls

235 lines (146 loc) · 12.2 KB

BioStructures.jl release notes

v4.4.2 - Mar 2025

  • A bug in storing indices for Transformation is fixed.

v4.4.1 - Mar 2025

  • A bug in input checking for Transformation is fixed.

v4.4.0 - Feb 2025

  • Selection strings now support interpolation.

v4.3.0 - Nov 2024

  • Selection strings are made much faster.

v4.2.1 - Oct 2024

  • Change compatibility bounds for new MetaGraphs.jl release.

v4.2.0 - Aug 2024

  • Base.copy is defined for structural elements and carries out a recursive copy of the element.
  • The fields of Transformation are made concrete, which may improve performance.

v4.1.1 - Aug 2024

  • PDB download functions now use HTTPS rather than FTP as the PDB will deprecate the FTP protocol.

v4.1.0 - Jul 2024

  • retrievepdb now takes the format keyword argument and uses MMCIFFormat by default.
  • MMTF files are no longer available to download via downloadpdb, downloadentirepdb, updatelocalpdb and downloadallobsoletepdb as the RCSB PDB no longer provides them.

v4.0.0 - Jun 2024

The package is made considerably more lightweight by moving a number of dependencies to extensions. This should make it easier for other packages to build on top of BioStructures.jl. Some types and functions are also renamed to avoid clashes, and a convenient string selection syntax is introduced.

Breaking changes

  • PDB, PDBXML, MMCIF and MMTF are renamed to PDBFormat, PDBXMLFormat, MMCIFFormat and MMTFFormat respectively to avoid clashing with module names. read(fp, PDB) should be replaced with read(fp, PDBFormat), for example.
  • ProteinStructure is renamed to MolecularStructure since it is not limited to representing protein structures.
  • x, y, z, x!, y! and z! are no longer exported as they are common variable names. They are still available as BioStructures.x etc.
  • Importing BioSequences.jl is now required to use LongAA.
  • Importing BioSequences.jl and BioAlignments.jl is now required to use pairalign, superimpose!, rmsd/displacements with the superimpose option or Transformation on structural elements.
  • Importing MMTF.jl is now required to use MMTFDict or writemmtf.
  • Importing DSSP_jll.jl is now required to use rundssp!, rundssp or the run_dssp option with read/retrievepdb.
  • Importing STRIDE_jll.jl is now required to use runstride!, runstride or the run_stride option with read/retrievepdb.

Non-breaking changes

  • Support for Julia versions before 1.9 is dropped.
  • A string selection syntax is introduced, allowing selections such as collectatoms(struc, sel"name CA and resnumber <= 5").
  • The selectors sidechainselector, proteinselector, acidicresselector, aliphaticresselector, aromaticresselector, basicresselector, chargedresselector, neutralresselector, hydrophobicresselector, polarresselector and nonpolarresselector are added.
  • PDB parsing in certain situations is now much faster.

v3.1.0 - May 2024

  • PrecompileTools.jl is used to reduce the time to first execution of PDB file reading.

v3.0.0 - Jan 2024

  • On Julia 1.9 and later the DataFrame and MetaGraph constructors are moved to package extensions in order to reduce the number of dependencies. Calling using DataFrames and using Graphs, MetaGraphs respectively is now required to access these functions.
  • The file formats PDB, PDBXML, MMCIF and MMTF are no longer subtypes of BioCore.IO.FileFormat, allowing BioCore.jl to be removed as a dependency.

v2.1.0 - Oct 2023

  • DSSP and STRIDE can now be run to assign secondary structure to proteins.

v2.0.0 - Feb 2023

  • The required versions of BioSequences.jl and BioAlignments.jl are updated to v3 of each, with support for earlier versions being dropped. LongAminoAcidSeq is hence renamed to LongAA, an alias for LongSequence{AminoAcidAlphabet}.
  • Fix bug in pdbentrylist.

v1.2.1 - Jan 2022

  • Fix bug allowing reflections during structural superimposition.

v1.2.0 - Dec 2021

  • firstindex and lastindex are defined for structural elements, contact maps and distance maps. This allows begin and end to be used in indexing expressions.
  • Support for Julia versions before 1.6 is dropped.

v1.1.0 - Nov 2021

  • The chainid! function is added, allowing the chain ID of a chain or residue to be changed. The new PDBConsistencyError is thrown when this would give an inconsistent structural state.
  • "WAT" is added to waterresnames and is hence used in waterselector and notwaterselector.
  • Switch from using LightGraphs.jl to using Graphs.jl.

v1.0.0 - May 2021

  • The ordering when sorting residues in a chain is changed from standard/hetero residue then residue number then insertion code to residue number then insertion code then standard/hetero residue. This makes in-chain hetero residues appear in the correct place in written PDB files.
  • Support for Julia versions before 1.3 is dropped.

v0.11.9 - Apr 2021

  • Fix bug in expanding disordered residues before applying residue selectors.
  • Change compatibility bounds for new DataFrames.jl release.

v0.11.8 - Mar 2021

  • Fix bug in expanding disordered atoms before applying atom selectors.

v0.11.7 - Dec 2020

  • Change compatibility bounds for new DataFrames.jl release.

v0.11.6 - Nov 2020

  • Change compatibility bounds for new Format.jl release.

v0.11.5 - Oct 2020

  • Some mmCIF files, such as the chemical component dictionary from the PDB, contain multiple data blocks. These can now be read in to a Dict{String, MMCIFDict} with readmultimmcif and written out with writemultimmcif.
  • Tab completion and an improved REPL display are added for MMCIFDict and MMTFDict.

v0.11.4 - Sep 2020

  • A ProteinStructure can now be obtained from a MMCIFDict or MMTFDict by passing them to the ProteinStructure constructor. This saves having to read the file twice when both the dictionary and the structure object are required.
  • Add get method for MMTFDict.

v0.11.3 - Sep 2020

  • Gzip support is added for reading and writing mmCIF files via the gzip keyword argument.

v0.11.2 - Sep 2020

  • Add get method for MMCIFDict.

v0.11.1 - Sep 2020

  • Fix bug in reading mmCIF data values containing a comment character.

v0.11.0 - Jun 2020

  • The required versions of BioSequences.jl and BioAlignments.jl are updated to v2 of each, with support for earlier versions being dropped. AminoAcidSequence is hence renamed to LongAminoAcidSeq.
  • threeletter_to_aa, a lookup table of amino acids, is re-exported from BioSymbols.

v0.10.1 - May 2020

  • Change compatibility bounds for new DataFrames.jl release.

v0.10.0 - Apr 2020

  • Change keyword argument names pdb_dir to dir and file_format to format for downloadpdb, downloadentirepdb, updatelocalpdb, downloadallobsoletepdb and retrievepdb.
  • Remove readpdb, which has the same functionality as read.
  • API reference section, more docstrings, links to related software and interactive Bio3DView.jl examples in documentation.

v0.9.4 - Apr 2020

  • Change compatibility bounds for new RecipesBase.jl and CodecZlib.jl releases.

v0.9.3 - Mar 2020

  • Change compatibility bounds for new RecipesBase.jl release.

v0.9.2 - Feb 2020

  • Improvements to performance throughout the package. Some functions are made up to 5 times faster.

v0.9.1 - Jan 2020

  • Fix documentation build.

v0.9.0 - Jan 2020

  • A reader and writer are added for the MMTF file format, building on top of MMTF.jl. The interface is the same as for PDB and mmCIF files, with files either being read into the standard hierarchical structure or a MMTFDict. Gzipped files are supported. PDB, mmCIF and MMTF files can be interconverted.
  • The expand_disordered flag is added to collectatoms, collectresidues, countatoms, countresidues, coordarray, writepdb, writemmcif, writemmtf and DataFrame. It determines whether disordered atoms and residues are expanded to include all entries. By default it is false except for the output functions, i.e. the last four above, where it is true by default.
  • The pdbextension dictionary is changed to remove leading dots in the values.
  • Improved file writing of empty elements.
  • Examples are split off into a separate section in the documentation.
  • A benchmark suite is added to track performance.

v0.8.0 - Dec 2019

  • Superimposition of structural elements is supported using the Kabsch algorithm. New functions are superimpose!, Transformation, applytransform! and applytransform.
  • rmsd and displacements carry out superimposition by default, with the relevant keyword arguments available. Setting superimpose to false prevents this. rmsdatoms and dispatoms respectively determine which atoms to calculate the property for.
  • The trivial allselector, which selects all atoms or residues, is added.
  • The backbone oxygen "O" is added to backboneatomnames.
  • Compatible bounds of package dependencies are added to Project.toml.

v0.7.0 - Oct 2019

  • MetaGraph from MetaGraphs.jl is extended to create graphs of contacting elements in a molecular structure, giving access to all the graph analysis tools in LightGraphs.jl.
  • DataFrame from DataFrames.jl is extended to allow creation of data frames from lists of atoms or residues.
  • pairalign from BioAlignments.jl is extended to produce pairwise alignments from structural elements.
  • AminoAcidSequence now takes any element type and has the gaps keyword argument.
  • Documentation example of interoperability with NearestNeighbors.jl.
  • Parametric types used more extensively internally.

v0.6.0 - Sep 2019

  • collectatoms, collectresidues, collectchains and collectmodels no longer run sort before returning the final list. The user can run an explicit sort themselves if desired. This change makes the functions faster and allows preservation of the element order.
  • Speed up residue iteration.
  • Documentation improvements.

v0.5.1 - Aug 2019

  • Fix MMCIFDict to always contain a Dict{String, Vector{String}} rather than a Dict{String, Union{String, Vector{String}}}, which includes making the "data_" tag a Vector{String}.
  • More functions documented and documentation bugfixes.

v0.5.0 - Aug 2019

  • The mmCIF reader now returns Array{String,1} for dictionary values even when there is only a single component. This improves consistency.
  • Documentation expanded with references to Bio3DView.jl and an extra example.
  • Replace REQUIRE with Project.toml.
  • Bugfix when reading truncated MODEL line in a PDB file.

v0.4.0 - Sep 2018

  • The ContactMap and DistanceMap types are introduced along with their supertype SpatialMap. contactmap is removed. Plot recipes are defined for visualisation of ContactMaps and DistanceMaps. showcontactmap provides a quick way to view a ContactMap in the terminal.
  • Bug fix on downloading MMTF files.

v0.3.0 - Aug 2018

  • Code is now compatible with Julia v0.7 and v1.0. Support for earlier Julia versions is dropped.
  • downloadpdb can now be given a function as the first argument, in which case the function is run with the downloaded filepath(s) as an argument and the file(s) are deleted afterwards.
  • Improved function docstrings.

v0.2.0 - Mar 2018

  • A reader and writer is added for the mmCIF format, which has been the standard PDB archive format since 2014. mmCIF files can either be read into a hierarchical structure object or directly in as a dictionary. PDB and mmCIF files can be interconverted.
  • chainid now returns a String instead of a Char. This allows multi-character chain IDs. This also changes chainids, chain and chains. Chains can be accessed by string (e.g. struc["A"]), but can still be accessed by character for single chain IDs (e.g. struc['A']).
  • show now returns a single line statement for objects across the module, in line with Julia conventions.

v0.1.0 - Aug 2017

Transfer of existing code from Bio.jl. Compatible with Julia v0.6.

Features:

  • Hierarchical data structure suitable for macromolecules, particularly proteins.
  • Fast reader and writer for the Protein Data Bank (PDB) file format.
  • Selection and iteration of structural elements.
  • Calculation of spatial properties such as distances and Ramachandran angles.
  • Functions to access the PDB.