release 0.11.3

Updated some documentation and made wheels for recent OS and pysam systems. Cython should NOT be required when using pip install except if you are compiling from source on purpose.

release 0.11.1

Bugfixes:

• htseq-count works with more than one additional attribute, just use --additional-attr more than once

Note that the change in how additional attributes are handled may break your unit tests or pipelines if they depend on htseq-count.

Have fun!

release 0.11.0

Bugfixes:

• default options for htseq-count are now --secondary-alignments ignore and --supplementary-alignments ignore
• linting of Cython files

Note that the change in default options may break your unit tests or pipelines if they depend on htseq-count. Turns out the new defaults are much more sensible in the majority of cases, because the mate of a secondary alignment is an ill-defined beast.

Have fun!

release 0.10.0

See PyPI for new releases: https://pypi.org/project/HTSeq/

release 0.9.1

Bugfix release for htseq-count:

• --secondary-alignments and --supplementary-alignments should now work for some corner cases of unmapped reads

release 0.9.0

This release adds a few options to htseq-count:

• --secondary-alignments handles secondary alignments coming from the same read
• --supplementary-alignments handles supplementary alignments (aka chimeric reads)

Raw but fast iterators for FASTA and FASTQ files have been added.

Support for the SAM CIGAR flags = and X (sequence match and mismatch) has been added.

Sequence objects can now be pickled/serialized.

Binaries for linux and OSX are now provided on PyPI.

Automation of the release process has been greatly extended, including OSX continuous integration builds.

Several bugs have been fixed, and some parts of the code have been linted or modernized.

release 0.8.0

This release adds a few options to htseq-count:

• --nonunique handles non-uniquely mapped reads
• --additional-attr adds an optional column to the output (typically for human-readable gene names)
• --max-reads-in-buffer allows increasing the buffer size when working with paired end, coordinate sorted files

Moreover, htseq-count can now take more than one input file and prints the output with one column per input file.

Finally, parts of the code have been streamlined or modernized, documentation has been moved to readthedocs, and other minor changes.

release 0.7.2

This release effectively merges the Python2 and Python3 branches.

Enhancements:

• pip install HTSeq works for both Python 2.7 and 3.4+

release_0.7.1py3

Version 0.7.01 Python 3 branch.

Enhancements:

• adapted for PyPI (direct installation of python3 version from PyPI not working yet).

release_0.7.1

Version 0.7.1, Python 2 (master) branch.

Enhancements:

• installs from PyPI