The updated Biopython Tutorial and Cookbook is online (PDF).

Platforms/Deployment

We currently support Python 2.5, 2.6 and 2.7 and also test under Python 3.1, 3.2 and 3.3 (including modules using NumPy), and Jython 2.5 and PyPy 1.9 (Jython and PyPy do not cover NumPy or our C extensions). We are still encouraging early adopters to help test on these platforms, and have included a ‘beta’ installer for Python 3.2 (and Python 3.3 to follow soon) under 32-bit Windows.

Please note we are phasing out support for Python 2.5. We will continue support for at least one further release (Biopython 1.62). This could be extended given feedback from our users. Focusing on Python 2.6 and 2.7 only will make writing Python 3 compatible code easier.

Features

GenomeDiagram has three new sigils (shapes to illustrate features). OCTO shows an octagonal shape, like the existing BOX sigil but with the corners cut off. JAGGY shows a box with jagged edges at the start and end, intended for things like NNNNN regions in draft genomes. Finally BIGARROW is like the existing ARROW sigil but is drawn straddling the axis. This is useful for drawing vertically compact figures where you do not have overlapping genes.

New module Bio.Graphics.ColorSpiral can generate colors along a spiral path through HSV color space. This can be used to make arbitrary ‘rainbow’ scales, for example to color features or cross-links on a GenomeDiagram figure.

The Bio.SeqIO module now supports reading sequences from PDB files in two different ways. The “pdb-atom” format determines the sequence as it appears in the structure based on the atom coordinate section of the file (via Bio.PDB,
so NumPy is currently required for this). Alternatively, you can use the “pdb-seqres” format to read the complete protein sequence as it is listed in the PDB header, if available.

The Bio.SeqUtils module how has a seq1 function to turn a sequence using three letter amino acid codes into one using the more common one letter codes. This acts as the inverse of the existing seq3 function.

The multiple-sequence-alignment object used by Bio.AlignIO etc now supports an annotation dictionary. Additional support for per-column annotation is planned, with addition and splicing to work like that for the SeqRecord per-letter annotation.

The Bio.Motif module has been updated and reorganized. To allow for a clean deprecation of the old code, the new motif code is stored in a new module Bio.motifs, and a PendingDeprecationWarning was added to Bio.Motif.

Experimental Code – SearchIO

This release also includes Bow’s Google Summer of Code work writing a unified parsing framework for NCBI BLAST (assorted formats including tabular and XML), HMMER, BLAT, and other sequence searching tools. This is currently available with the new BiopythonExperimentalWarning to indicate that this is still somewhat experimental. We’re bundling it with the main release to get more public feedback, but with the big warning that the API is likely to change. In fact, even the current name of Bio.SearchIO may change since unless you are familiar with BioPerl its purpose isn’t immediately clear.

Contributors

• Brandon Invergo
• Bryan Lunt (first contribution)
• Christian Brueffer (first contribution)
• David Cain
• Eric Talevich
• Grace Yeo (first contribution)
• Jeffrey Chang
• Jingping Li (first contribution)
• Kai Blin (first contribution)
• Leighton Pritchard
• Lenna Peterson
• Lucas Sinclair (first contribution)
• Michiel de Hoon
• Nick Semenkovich (first contribution)
• Peter Cock
• Robert Ernst (first contribution)
• Tiago Antao
• Wibowo ‘Bow’ Arindrarto

• BioRuby
• Biopython
• BioPerl

The BioRuby team are also using Travis-CI for automated testing of their new ‘plugin’ ecosystem, BioRuby Gems, or BioGems.

Travis-CI gives us continuous testing, but for the moment only covers one operating system (currently 32 bit Ubuntu Linux using Virtual Machines). This automated testing is therefore complementary to our existing OBF BuildBot server which aims to run nightly tests on volunteer developer machines setup to cover a broad range of operating systems and configurations.

However, Travis-CI are working on a new feature – automatic testing of pull requests, currently only available on a donation basis – which the OBF was happy to support.

What this means is that when a contributor has some code ready for integration, they can issue a GitHub pull request, and then Travis-CI will automatically run the unit tests with those proposed changes. This is something that currently the core-developers would normally do manually as part of evaluating proposed changes, so having this happen automatically should be a big help.

We’re excited about making more use of Travis-CI for other OBF projects. Thus far we’ve been really impressed with Travis-CI.

Platforms/Deployment

We currently support Python 2.5, 2.6 and 2.7 and also test under Jython 2.5 and PyPy 1.9 (which does not cover NumPy or our C extensions). Please note that Python 2.4 or earlier is not supported.

Most functionality is also working under Python 3.1 and 3.2 (including modules using NumPy). We are now encouraging early adopters to help beta testing on these platforms, and have included a ‘beta’ installer for Python 3.2 under 32-bit Windows.

Features

New module Bio.bgzf supports reading and writing BGZF files (Blocked GNU Zip Format), a variant of GZIP with efficient random access, most commonly used as part of the BAM file format and in tabix. This uses Python’s zlib library internally, and provides a simple interface like Python’s gzip library. Using this the Bio.SeqIO indexing functions now support BGZF compressed sequence files. See this blog post for more background on BGZF and SeqIO.

The GenBank/EMBL parser will now give a warning on unrecognized feature locations and continue parsing (leaving the feature’s location as None). Previously it would abort with an exception, which was often unhelpful.

The Bio.PDB.MMCIFParser is now compiled by default (but is still not available under Jython, PyPy or Python 3).

The SFF parser in Bio.SeqIO now decodes Roche 454 ‘universal accession number’ 14 character read names, which encode the timestamp of the run, the region the read came from, and the location of the well.

In the Phylo module, the “draw” function for plotting tree objects has become much more flexible, with improved support for matplotlib conventions and new parameters for specifying branch and taxon labels. Writing in the PhyloXML format has been updated to more closely match the output of other programs. A wrapper for the program RAxML has been added under Bio.Phylo.Applications, alongside the existing wrapper for PhyML.

Additionally there have been other minor bug fixes and more unit tests.

Contributors

Many thanks to the Biopython developers and community for making this release
possible, especially the following contributors:

• Brandon Invergo
• Eric Talevich
• Jeff Hussmann (first contribution)
• John Comeau (first contribution)
• Kamil Slowikowski (first contribution)
• Kevin Jacobs
• Lenna Peterson (first contribution)
• Matt Fenwick (first contribution)
• Peter Cock
• Paul T. Bathen
• Wibowo Arindrarto

I’m very pleased and excited to announce that the Open Bioinformatics Foundation has selected 5 very capable students to work on OBF projects this summer as part of the Google Summer of Code (GSoC) program.

The accepted students, their projects, and their mentors (in alphabetical order):

• Wibowo Arindrarto:
  SearchIO Implementation in Biopython
  mentored by Peter Cock
• Lenna Peterson:
  Diff My DNA: Development of a Genomic Variant Toolkit for Biopython
  mentored by Brad Chapman, Reece Hart, James Casbon
• Marjan Povolni:
  The worlds fastest parallelized GFF3/GTF parser in D, and an interfacing biogem plugin for Ruby
  mentored by Pjotr Prins, Francesco Strozzi, Raoul Bonnal
• Artem Tarasov:
  Fast parallelized GFF3/GTF parser in C++, with Ruby FFI bindings
  mentored by Pjotr Prins, Francesco Strozzi, Raoul Bonnal
• Clayton Wheeler:
  Multiple Alignment Format parser for BioRuby
  mentored by Francesco Strozzi and Raoul Bonnal

As in every year, we received many great applications and ideas. However, funding and mentor resources are limited, and we were not able to accept as many as we would have liked. Our deepest thanks to all the students who applied: we sincerely appreciate the time and effort you put into your applications, and hope you will still consider being a part of the OBF’s open source projects, even without Google funding. I speak for myself and all of the mentors who read and scored applications when I say that we were truly honored by the number and quality of the applications we received.

For the accepted students: congratulations! You have risen to the top of a very competitive application process. Now it’s time to “put your money where your mouth is”, as the saying goes. Let’s get out there and write some great code this summer!

Best regards,

Robert Buels
OBF GSoC 2012 Organization Administrator

This example attempts a simplified reproduction of Figure 6 in Proux et al. (2002), and shows three related phage genomes one above the other. Different classes of genes have been given different colors, while the strength of the red shaded cross-links indicates the percentage identity of the linked genes. Note there are some minor differences in the GenBank annotation we’ve used and the genes shown in the original figure.

Note while this example is in the Tutorial HTML and PDF online, it was not in the zip/tarball for Biopython 1.59, and nor was the Proux_et_al_2002_Figure_6.py script. It will be in future releases.

Another motivating use case for this functionality was to produce vector images of whole genome alignments in the style of the Artemis Comparison Tool (ACT) or Mauve. We’ve got a poster printer in the building just crying out to be used for showing whole genome comparison of a dozen bacteria strains!

Platforms/Deployment

We currently support Python 2.5, 2.6 and 2.7 and also test under Jython 2.5 (which does not cover NumPy). Please note that this release will not work on Python 2.4

Most functionality is also working under Python 3.1 and 3.2 (including modules using NumPy), and under PyPy (excluding our NumPy dependencies). We are now encouraging early adopters to help beta testing on these platforms.

The installation setup.py now supports ‘install_requires’ when setuptools is installed. This avoids the manual dialog when installing Biopython via easy_install or pip and numpy is not installed. It also allows user libraries that require Biopython to include it in their install_requires and get automatic installation of dependencies.

Features

New module Bio.TogoWS offers a wrapper for the TogoWS REST API, a web service based in Japan offering access to KEGG, DDBJ, PDBj, CBRC plus access to some NCBI and EBI resources including PubMed, GenBank and UniProt. This is much easier to use than the NCBI Entrez API, but should be especially useful for Biopython users based in Asia.

The NCBI Entrez Fetch function Bio.Entrez.efetch has been updated to handle the NCBI’s stricter handling of multiple ID arguments in EFetch 2.0 (released February 2012, see this announcement), however the NCBI have also changed the retmode default argument so you may need to make this explicit. e.g. add retmode="text" to your EFetch calls (see this announcement).

The position objects used in Bio.SeqFeature now act almost like integers, making dealing with fuzzy locations in EMBL/GenBank files much easier. Also the SeqFeature‘s strand and any database reference are now properties of the FeatureLocation object (a more logical placement), with proxy methods for backwards compatibility.

Bio.Graphics.BasicChromosome has been extended to allow simple sub-features to be drawn on chromosome segments, suitable to show the position of genes, SNPs or other loci.

Bio.Graphics.GenomeDiagram has been extended to allow cross-links between tracks, and track specific start/end positions for showing regions. This can be used to imitate the output from the Artemis Comparison Tool (ACT). Also, a new attribute circle_core makes it easier to have an empty space in the middle of a circular diagram (see tutorial).

Note Bio.Graphics requires the ReportLab library.

Bio.Align.Applications now includes a wrapper for command line tool Clustal Omega for protein multiple sequence alignment.

Bio.AlignIO now supports sequential PHYLIP files (as well as interlaced PHYLIP files) as a separate format variant.

Additionally there have been other minor bug fixes and more unit tests, and updates to the documentation including the Biopython Tutorial (PDF).

Contributors

Many thanks to the Biopython developers and community for making this release possible, especially the following contributors:

• Andreas Wilm (first contribution)
• Alessio Papini (first contribution)
• Brad Chapman
• Brandon Invergo
• Connor McCoy
• Eric Talevich
• João Rodrigues
• Konrad Förstner (first contribution)
• Michiel de Hoon
• Matej Repič (first contribution)
• Leighton Pritchard
• Peter Cock

Here’s a PDF version too. This example just parses the Arabidopsis thaliana GenBank files to get the chromosome lengths and the tRNA gene placements. There are so many tRNA on the forward strand of Chr I that their labels are forced to overlap. Here the figure just uses a different color for each chromosome, but you can color each feature individually.

This kind of diagram is often used for showing the placement of SNPs or other loci of interest (e.g. disease or breeding markers).

P.S. Biopython produces these and other graphics using ReportLab, an open source Python library capable of outputting PDF, SVG, PNG, etc. Very handy.

A new interface and parsers for the PAML (Phylogenetic Analysis by Maximum Likelihood) package of programs, supporting codeml, baseml and yn00 as well as a Python re-implementation of chi2 was added as the Bio.Phylo.PAML module.

Bio.SeqIO now includes read and write support for the SeqXML, a simple XML format offering basic annotation support. See Schmitt et al (2011) in Briefings in Bioinformatics.

Bio.SeqIO now includes read support for ABI files (“Sanger” capillary sequencing trace files, containing called sequence with PHRED qualities).

The Bio.AlignIO “fasta-m10″ parser was updated to cope with the marker lines as used in Bill Pearson’s FASTA version 3.36, without this fix the parser would only return alignments for the first query sequence.

The Bio.AlignIO “phylip” parser and writer now treat a dot/period in the sequence as an error, in line with the official PHYLIP specification. Older verions of our code didn’t do anything special with this character. Also, support for “phylip-relaxed” has been added which allows longer record names as used in RAxML and PHYML.

Of potential interest to anyone subclassing Biopython objects, any remaining “old syle” Python classes have been switched to “new style” classes. This allows things like defining properties.

Bio.HMM’s Viterbi algorithm now expects the initial probabilities explicitly.

Many thanks to the Biopython developers and community for making this release possible, especially the following contributors:

• Aaron Gallagher (first contribution)
• Bartek Wilczynski
• Bogdan T. (first contribution)
• Brandon Invergo (first contribution)
• Connor McCoy (first contribution)
• David Cain (first contribution)
• Eric Talevich
• Fábio Madeira (first contribution)
• Hongbo Zhu
• Joao Rodrigues
• Michiel de Hoon
• Peter Cock
• Thomas Schmitt (first contribution)
• Tiago Antao
• Walter Gillett
• Wibowo Arindrarto (first contribution)

Bio.SeqIO now includes an index_db() function which extends the existing
indexing functionality to allow indexing many files, and more importantly
this keeps the index on disk in a simple SQLite3 database rather than in
memory in a Python dictionary.

Bio.Blast.Applications now includes a wrapper for the BLAST+ blast_formatter
tool from NCBI BLAST 2.2.24+ or later. This release of BLAST+ added the
ability to run the BLAST tools and save the output as ASN.1 format, and then
convert this to any other supported BLAST ouput format (plain text, tabular,
XML, or HTML) with the blast_formatter tool. The wrappers were also updated
to include new arguments added in BLAST 2.2.25+ such as -db_hard_mask.

The SeqRecord object now has a reverse_complement method (similar to that of
the Seq object). This is most useful to reversing per-letter-annotation (such
as quality scores from FASTQ) or features (such as annotation from GenBank).

Bio.SeqIO.write’s QUAL output has been sped up, and Bio.SeqIO.convert now
uses an optimised routine for FASTQ to QUAL making this much faster.

Biopython can now be installed with pip. Thanks to David Koppstein and
James Casbon for reporting the problem.

Bio.SeqIO.write now uses lower case for the sequence for GenBank, EMBL and
IMGT output.

The Bio.PDB module received several fixes and improvements, including starting
to merge João’s work from GSoC 2010; consequently Atom objects now know
their element type and IUPAC mass. (The new features that use these
attributes won’t be included in Biopython until the next release, though, so
stay tuned.)

The nodetype hierachy in the Bio.SCOP.Cla.Record class is now a dictionary
(previously it was a list of key,value tuples) to better match the standard.

Many thanks to the Biopython developers and community for making this release
possible, especially the following contributors:

• Brad Chapman
• Eric Talevich
• Erick Matsen (first contribution)
• Hongbo Zhu
• Jeffrey Finkelstein (first contribution)
• Joanna & Dominik Kasprzak (first contribution)
• Joao Rodrigues
• Kristian Rother
• Leighton Pritchard
• Michiel de Hoon
• Peter Cock
• Peter Thorpe (first contribution)
• Phillip Garland
• Walter Gillett (first contribution)

Feedback is most welcome on the mailing lists, or redmine.

GSoC is a Google-sponsored student internship program for open-source projects, open to students from around the world (not just US residents). Students are paid a $5000 USD stipend to work as a developer on an open-source project for the summer. For more on GSoC, see GSoC 2011 FAQ.

Student applications are due April 8, 2011 at 19:00 UTC. Students who are interested in participating should look at the OBF’s GSoC page, which lists project ideas, and whom to contact about applying.

For current developers on OBF projects, please consider volunteering to be a mentor if you have not already, and contribute project ideas. Just list your name and project ideas on the OBF wiki page, and on the relevant project’s GSoC wiki page.

Thanks to all who helped make OBF’s application to GSoC a success, and let’s have a great, productive summer of code!

Rob Buels
OBF GSoC 2011 Administrator