- Added environment.yaml for dependency version reproducibility. (New version of ChangeO MakeDb.py has new commands and struggles parsing IgBlast 1.10.0 output)
- General cleanup
- New fasta parser that can handle multi-line records
Most changes pertain to the short pipeline. Moved main scripts into bin folder.
- The sequence beyond anchor is no longer removed from output reads.
- J end UMI moved from deduplication, so that all UMI extraction is now preformed at precleaning stage.
- Fixed comments in R script that caused J germline plots not to be produced.
- Check if germline plot is produced.
- Fixed an incorrect mispriming correction offset for the kappa locus.
- Extended error estimation for human sequences.
- json file written with identity of reads assembled and unassembled to allow merging after deduplication.
- Option to output sequences with ambiguous N nucleotides (normally these are filled in with basepairs from sequences best matching the consensus).
- Fixed bug in UMI report tables (most likely cause by pandas update).
- Simplified options, made MSA available to all pipelines.
- Moved general functions into deduplicattion_general.py and split deduplicate_bundle_parallel for clarity.
- Updated UMI correction to work with latest version.
- Added json option so output reads can be marked assembled or unassembled.
- DJ reads are also filtered by the V end IgBlast calls, instead on just J end.
- Moved package data into MANIFEST.in to enable
pip --editablemode to work with pkg_resources.resource_filename.
- Initial release