Log message:
*: limit Python versions due to py-numpy dropping Python 3.11 support

Log message:
py-biopython: update to 1.87.

30 March 2026: Biopython 1.87
=============================

Migrated from ``setup.py`` to ``pyproject.toml`` for packaging configuration.

Addressed security issue CVE-2025-68463 in ``Bio.Entrez.Parser`` if parsing
untrusted files.

Additionally, a number of small bugs and typos have been fixed with additions
to the test suite and type annotations.

28 October 2025: Biopython 1.86
===============================

``Bio.SearchIO`` now supports parsing the tabular and plain text output of
`Infernal <http://eddylab.org/infernal/>` (v1.0.0+) RNA search tool. The
format names are ``infernal-tab`` and ``infernal-text``.

The default value of the gap score of a ``PairwiseAligner`` object was changed
in this release.  Previously, for consistency with ``Bio.pairwise2``, the
default value for gap score was 0.  However, this means that a mismatch, an
insertion followed by a deletion, and a deletion followed by an insertion all
get assigned a score of 0.  The aligner then finds a large number of alignments
that are logically the same, but have trivial differences between them.  For
example, aligning AAACAAA to AAAGAAA previously yielded the following three
alignments, all with score 6::

     AAACAAA        AAAC-AAA        AAA-CAAA
     AAAGAAA        AAA-GAAA        AAAG-AAA

With the new default parameter for the gap score, only the first alignment is
returned.

``Bio.PDB.PDBIO`` now ensures that b-factor values are always at most 6 characters to
ensure that we do not violate the wwPDB specification. This should not have an impact
on the majority of uses, as b-factor values are generally small (less than 100). When
1000 \<= b-factor \< 10_000, the value is rounded to a single decimal \ 
place. When,
10_000 \<= b-factor \< 999_999, the value is rounded to zero decimal \ 
places. Values
above 999_999 are now clamped. The justification for this is the rise in the b-factor
field being used for additional metadata, typically from computational tools.

``Bio.Align`` now provides a method ``Alignment.from_alignments_with_same_reference``
to construct a multiple sequence alignment from a collection of alignments
that share the same reference sequence.

``Bio.PDB.PDBIO`` will now raise module specific warnings:
``Bio.PDB.PDBExceptions.PDBIOWarning``.

``Bio.PDB.SCADIO`` now supports object selection by color in the OpenSCAD
output file.  This enables generation of separate STL files for each color for
printing protein structures on multi-material 3D printers.

The ``iplotx`` library is mentioned in the Tutorial as an option to visualise
trees using complex style options.

Log message:
*: remove reference to (removed) Python 3.9

Log message:
*: py-numpy needs Python >= 3.11 now

Log message:
Fix PLIST after py-setuptools update; bump depends and revision

Log message:
py-biopython: updated to 1.85

Biopython 1.85

This release of Biopython supports Python 3.9, 3.10, 3.11, 3.12 and 3.13. It
has also been tested on PyPy3.9 v7.3.13. Python 3.9 is approaching end of
life, our support for it is now deprecated.

Some optimisation work was done for ``Bio.SeqIO`` including avoiding nested
iterators, and speeding up both FASTA and FASTQ parsing.

``Bio.motifs`` now supports reading PFM from Cys2His2 Zinc Finger Proteins PWM
Predictor and reading motifs in ``pfm-four-columns`` format will set motif name
to "" instead of None, when no motif name was found.

Tests that use assertAlmostEqual calls now use ``places`` parameter with enough
presision when comparing very small numbers in scientific notation.

``Bio.motifs`` now supports reverse complementing RNA motifs and correctly
generating degenerate consensus sequences for RNA motifs.

``Bio.motifs.minimal`` now supports parsing RNA motifs and parsing motifs
for which not all statistics are provided (e.g. missing E-values or nsites).

``Bio.motifs.clusterbuster`` now supports parsing GAP and WEIGHT parameters
and can optionally write Cluster Buster motif files with floats instead of
integers, by specifying the ``precision=<int>`` parameter when writing:
e.g. motifs.write(motifs, "clusterbuster", precision=2)

Additionally, a number of small bugs and typos have been fixed with additions
to the test suite and type annotations.

Log message:
py-*: remove unused tool dependency

py-setuptools includes the py-wheel functionality nowadays

Log message:
*: clean-up after python38 removal