v1.5.3

• Fixed memory limit in Muscle library so that multiprocessing doesn't crash
• Changed parsnp compilation to -O2 to fix WSL2 bug
• Added warning for trying to use too many threads

Release v1.5.2

• Fixed fallback to FastTree on RaxML failure
• Made --curate include all genomes regardless of length

Release v1.5.1

Bugfixes for Parsnp v1.5.0:

• Mash and FasANI recruitment now works regardless of reference type
• Phiprofile fixes for python 3
• Logging input errors is now cleaner

Release v1.5.0

• Ported to Python 3, no longer supports Python 2
• Added pythonic logging and argument parsing
• Merged build instructions for macOS and Unix
• Switched FastTree to RaxML
• All dependency tools must be on the system path (no longer defaults to bin/)
• More robust error checking (skips bad files instead of crashing, more warnings + debug mode)
• Fixed .gbk parsing
• Convenient input now allows sequences to be supplied via regex i.e. what use to be parsnp -d ref/ -r ! can now be written as parsnp -d ref/*.fna which allows for heterogeneous input directories and spread out data.

Test release of v1.5

v1.5beta

update tag

v1.2

Bug fixes

• Important fix to FastTree2 binary to enable use of double precision branch lengths (instead of minimum length of 0.0001); FastTree2 issue fully detailed in the following blog post: http://darlinglab.org/blog/2015/03/23/not-so-fast-fasttree.html.

v1.1.1

Fix to fully disable/deprecate -R parameter

v1.1

Features

• Added support to output unaligned regions in XMFA format (--unaligned). This output format contains a list of all unaligned sequence due to: (i) unrelated sequence, (ii) failure to extend past match (MUM) boundaries/sensitivity limitations, and (iii) subset relationships. This output file is further described in the documentation --> http://harvest.readthedocs.org/en/latest/.
• Updated MUMi parameter to allow for setting a maximum MUMi value to enable bypassing the distribution cutoff

Bug fixes

• Fixed minor bug when parsing multiple GenBank files for harvesttools annotation import
• Updated branch lengths via harvest-tools
• Fixed bug that produced VCF output without repeat filtering, even if enabled
• Fixed issue where empty fasta file was causing index errors
• Fixed exit code to be 0 on –h (Issue #4)
• Updated docs to indicate CXXFLAGS=’-fopenmp’ must be provided to configure for openmp support
• Added support for harvesttools v1.2 (see: https://github.com/marbl/harvest-tools/releases/tag/v1.2)

Changes

• Added –V, --version parameter
• Allow columns with Ns to be marked as pass in VCF
• Allow any fasta file present in -d, regardless of extension, to be used
• Disallow genome file names containing spaces or the following special characters: . , [ ] { } ( ) ! ; " ' * ? < > |
• Default repeat filtering disabled, removing NUCmer dependency. Repeat filtering still possible via bed format support and harvest-tools; see harvesttools documentation for further details.

v1.0

NEWS

07/21/2014: First official release

DOCS

For parsnp user guide please see:

http://harvest.readthedocs.org/en/latest/content/parsnp.html

MD5 checksums

08c19b4d12e8199b2ce098550b4100e6 parsnp-OSX64-v1.0.tar.gz
f82b6b9dae456fe9263ee6214b2633af parsnp-Linux64-v1.0.tar.gz