We are pleased to announce the release of Biopython 1.49. There have been some significant changes since Biopython 1.48 was released a few months ago, which is why we initially released a beta for wider testing. Thank you to all those who tried this and reported the minor problems uncovered.
As previously announced, the big news is that Biopython now uses NumPy rather than its precursor Numeric (the original Numerical Python library).
As in the previous releases, Biopython 1.49 supports Python 2.3, 2.4 and 2.5 but should now also work fine on Python 2.6. Please note that we intend to drop support for Python 2.3 in a couple of releases time (see previous news article).
We also have some new functionality, starting with the basic sequence object (the Seq class) which now has more methods. This encourages a more object orientated coding style, and makes basic biological operations like transcription and translation more accessible and discoverable.
Our BioSQL interface can now optionally fetch the NCBI taxonomy on demand when loading sequences (via Bio.Entrez) allowing you to populate the taxon/taxon_name tables gradually. Also, BioSQL should now work with the psycopg2 driver for PostgreSQL (as well as the older psycopg driver), and the handling of feature locations has also been improved.
Finally, our old parsing infrastructure (Martel and Bio.Mindy) is now considered to be deprecated, meaning mxTextTools is no longer required to use Biopython. This should not affect any of the typically used parsers (e.g. Bio.SeqIO and Bio.AlignIO).
Given there have been more changes than in recent Biopython releases, please do check your old scripts still work fine, and let us know on the mailing list or file a bug if there is anything wrong.
Source distributions and Windows installers are available from the Biopython website: biopython.org
-Peter on behalf of the Biopython developers