Biotite

Fixes

- Support for new UniProt REST API (409)
- Preserve lower-case chain IDs when an `AtomArray` is read from PDB and PDBQT files (413)
- `application.vina.VinaApp` supports now docking of molecules containing certain metal elements

Changelog

Additions

- Support for new RCSB search API (408)
- Added `case_sensitive` parameter in `database.rcsb.FieldQuery`
- `structure.info.mass()` support deuterium
- `structure.connect_via_distances()` can connect atoms over periodic boundaries
- Added more chain-level utilities consistent with residue-level utilities
- `structure.apply_chain_wise()`
- `structure.spread_chain_wise()`
- `structure.get_chain_masks()`
- `structure.get_chain_starts_for()`
- `structure.get_chain_positions()`
- `structure.superimpose()` supports also pure coordinates


Changes

- `structure.hbond()` uses an associated `structure.BondList` to find hydrogen atoms to potential hydrogen bond donors
- Lines depicting bonds in `structure.graphics.plot_atoms()` and `structure.graphics.plot_ball_and_stick_model()` use rounded tips


Fixes

- Fixed `structure.io.pdbx.get_assembly` missing chains in some structures (387)
- Added a more meaningful error, if *Matplotlib* is required, but not installed (302)
- Added more descriptive error, if a `structure.io.pdb.PDBFile` has erroneous atom IDs (379)
- `structure.io.pdb.PDBFile` pads lines always to 80 characters
- Allow empty attribute string in `sequence.io.GFFFile`
- Fixed wrong similarity scores, if a `sequence.align.SubstitutionMatrix` with two different alphabets is read from string or file
- Fixed `application.mafft.MafftApp` runs for more than 10 sequences.

Changelog

Additions

- Added `application.muscle.Muscle5App` to support the changed CLI of *Muscle 5*
- Added `structure.orient_principal_components()` to orient atom coordinates
to the given axes
- `biotite.structure.io.pdbx.get_structure()` uses `label_xxx` or `auth_xxx`
field as fallback, if the respective other one is not available
- Added `default_bond_type` parameter to
`biotite.structure.io.write_structure_to_ctab()` and
`biotite.structure.connect_via_distances` to allow the user to change the
`BondType` in the generated `BondList`


Fixes

- `sequence.io.gff.GFFFile.read()` is now able to read GFF records with trailing
tabs
- Fixed `DeprecationWarning` in `structure.align_vectors()` (295)
- Fixed alignment in atom name column in
`structure.io.pdb.PDBFile.write()`
- Fixed error handling in `structure.index_xxx()` functions, if invalid input
shape is given
- Ensured quoted values in looped categories will not be truncated in
`structure.io.pdbx.PDBxFile.set_category()`

Changelog

Additions

- Add Python 3.10 build
- Add `writer_iter()` to some `File` classes
- Allows entry-wise writing into a file to reduce memory consumption, as
there is no longer the need to keep the entire file content in memory
- Implemented for `sequence.io.FastaFile`, `sequence.io.FastqFile` and all
classes inheriting from `structure.io.TrajectoryFile`
- Add `symbol_spacing` parameter to `sequence.graphics.plot_alignment`
- If the parameter is given, a small gap is introduced every *n*
symbols in the alignment
- Also added to `sequence.graphics.plot_alignment_similarity_based` and
`sequence.graphics.plot_alignment_types_based`


Changes

- Python 3.7 builds are discontinued

Fixes

- Now the simulation time for each frame is returned in
`structure.io.NetCDFFile`, previously it was always `None`

Changelog

Additions
- New functionalities for `sequence.SequenceProfile`
- Added method `probability_matrix()` to compute the symbol probabilities
from total frequencies
- Added method `log_odds_matrix()` to calculate a position weight matrix
- Added methods `sequence_probability()` and `sequence_score()` to assess the
adherence of a given sequence to a profile
- New functionalities for `structure.BondList`
- New bond type `structure.BondType.AROMATIC_TRIPLE` to support triple bonds
in aromatic systems
- Added `structure.BondList.without_aromaticity()` to convert bonds with
`structure.BondType.AROMATIC_<order>` to `structure.BondType.<order>`
- Added `structure.info.bond_order()`
- Used to get the `structure.BondType` of the bond between two atoms in a
residue
- Replaces `structure.info.bond_order()`
- Initial loading of bond dataset is much faster


Changes
- Deprecated `structure.info.bond_order()`


Fixes
- Fixed `structure.io.pdb.PDBFile.get_structure()` raising an exception,
if the PDB file contains an invalid `CRYST1` record, now a warning is printed
- Fixed `CONECT` records written by `structure.io.pdb.PDBFile.set_structure()`
- Previously, an empty second `CONECT` record was created, if a atom has 4
bond partners
- Fixed installation of Biotite source distribution with Python 3.10 (356)
- Fixed running `application.muscle.MuscleApp` with nucleotide sequences

Changelog


Changes
- `sequence.graphics.plot_sequence_logo()` requires now a
`sequence.SequenceProfile`, usage of `sequence.align.Alignment` still works,
but is deprecated
- `database.rcsb.FieldQuery` has now an optional `molecular_definition`
parameter to allow searches in molecule related fields.


Fixes
- The new RCSB search API is now supported (347)
- More than two `database.rcsb.Query` objects can be combined with logical
operators
- `application.blast.BlastWebApp` now also reports hit sequences containing
selenocysteine without errors (344)
- `Atom`, `AtomArray` and `AtomArrayStack` can be unpickled (349)
- `sequence.graphics.plot_sequence_logo()` supports now all Matplotlib backends
(345)