GitLab

Current version: 3.0.1b4

You can check the version installed by running obi --version

How to update with pip:

From the OBITools3 virtual environment, run:

pip3 install --pre --upgrade OBITools3

How to update with git:

From the OBITools3 installation directory and virtual environment, run:

git pull
python3 setup.py install

Release notes:

• 3.0.1b4:

  • ngsfilter: fixed critical bug where barcodes shorter than the forward primer would be missed

  • import: fixed the import of tabular files with no header

• 3.0.1b3:

  • uniq: now OK to use -m option even if only one unique key in information to merge (e.g. one sample)

• 3.0.1b2:

  • Made the OBITools3 more 'empty file friendly'

• 3.0.1b1:

  • The OBITools3 now work on the Windows Subsystem for Linux

• 3.0.0b43:

  • Dictionaries with one key are now handled

• 3.0.0b42:

  • import: improved Genbank file parser

• 3.0.0b41:

  • less: fixed an issue when trying to display 'tabular data' such as ngsfilter files

• 3.0.0b40:

  • Fixed compilation issue with latest versions of gcc

• 3.0.0b39:

  • export: fixed a bug where exporting to tab format with a header would not export the first line of data

• 3.0.0b38:

  • grep: fixed --id-list option

• 3.0.0b37:

  • import: can now import SILVA fasta files, and now handling sequences with Uracil (U) nucleotides (by converting them to Thymine (T) for now)

  • addtaxids: new command to annotate sequences with their NCBI taxid based on their taxon name

• 3.0.0b36:

  • ls: reworked a bit (mostly just now done in C instead of Cython)

• 3.0.0b35:

  • ecotag: added separate threshold for minimum circle identity (sequence is assigned to the LCA of all sequences within a similarity circle of the best matches; the threshold for this circle is the highest value between the chosen minimum circle identity and the best assignment score found for the query sequence)

• 3.0.0b34:

  • cat: fixed open file descriptor leak

• 3.0.0b33:

  • cat: fixed a bug introduced in 3.0.0b28 (KeyError: 'noprogressbar')

• 3.0.0b32:

  • export: now automatically sorts dictionary keys alphabetically for tab/csv output
  • various small bug fixes and improvements

• 3.0.0b31:

  • export: fixed a bug introduced in version v3.0.0b28
  • ls: added more information about view creation date and OBITools3 version

• 3.0.0b30: alignpairedend: fixed division by 0 error introduced in version v3.0.0b28

• 3.0.0b29:

  • import: added --space-priority option to prioritize saving space over time when importing a view with a line selection associated
  • fixed and improved the associated column system

• 3.0.0b28:

  • all commands: now handle output to standard output (use - as output URI)
  • ngsfilter: unidentified sequences are now kept untrimmed
  • alignpairedend: improved/fixed the output tags for the alignment score and lengths
  • grep: now prints the number of entries grepped

• 3.0.0b27:

  • uniq: fixed a bug where some sequences were not properly dereplicated
  • clean: fixed the behaviour when no or bad sample information is given

• 3.0.0b26:

  • export: dictionaries are now formatted as they were in the original OBITools, and fixed tuple formatting
  • ngsfilter: reversed tag is now set to 'False' instead of 'None' when false.

• 3.0.0b25:

  • ls: fixed an issue with big DMS
  • ngsfilter: made one of the tag error messages more accurate ('No tags found' changed to 'No sample with that tag combination')
  • grep and uniq now output empty views when the input is an empty view

• 3.0.0b24:

  • Improved ecotag output with list of taxids of best matches
  • Improved uniq progress bar
  • Fixed bash history display for views
  • Fixed bugs with the output of sequences unaligned ('joined') by alignpairedend
  • Fixed a bug where tuple columns would 'lose' their indexer in some cases

• 3.0.0b23: Fixed installation on Ubuntu without pip (problem introduced in 3.0.0b20).

• 3.0.0b22: Fixed a bug when skipping unreadable sequences in obi import.

• 3.0.0b21: Improved import of Genbank seq files.

• 3.0.0b20: Reworked the installation (and version formatting) to work with pip, and made obi import work with seq Genbank extension (+ other minor import fixes).

• 3.0.0-beta17: Fixed a bug in obi import stopping the program when the last line of a fasta file contained a single nucleotide.

• 3.0.0-beta16: Fixed the import and dereplication of old obitools files containing premerged informations, and other minor improvements.

• 3.0.0-beta15: Various fixes and improvements, including notable bugs in ecotag and build_ref_db that made ecotag less specific than it should have been.

• 3.0.0-beta14:

  • Fixed a memory bug when importing a taxdump and building the list of merged taxids
  • Improved disk space management when filling columns

• 3.0.0-beta13:

  • obi import: fixed a critical bug where the last entry of embl and genbank files was not imported.

  • ecopcr now prints a warning instead of stopping when not finding a taxid.

  • Various other small bug fixes, and improvements in documentation, error messages and checks.

• 3.0.0-beta11: Fixed a bug in obi cat when concatenating views with columns having different key sets, and a bug in obi ecotag generating badly formatted column comments.

• 3.0.0-beta10: obi import: new option --preread to do a first readthrough of the dataset if it contains huge dictionaries for a much faster import.

• 3.0.0-beta9: Better disk space management when importing files (rewritten columns are deleted).

• 3.0.0-beta8: Fixed a bug when importing a file where the first entry would contain a dictionary with a single key.

• 3.0.0-beta7: Added the 'obi' prefix in the bash history commands, and fixed a compilation issue with some versions of gcc.

• 3.0.0-beta6: New command: obi cat to concatenate views, and various inconsequential fixes.

• 3.0.0-beta5: Fixed a bug in ngsfilter when there is only one tag.

• 3.0.0-beta4: Fixed and improved the options to keep nucleotides around the amplicon in ecopcr.

• 3.0.0-beta3: Fixed bugs in ngsfilter when cutting sequences, and a few minor bugs.