TranscriptClean v3.0

TranscriptClean and its dependencies are now installable via pip and runnable from the command line without providing a path to the Python file. See README.md for details on how to run now.

TranscriptClean v2.1

Making a release to package TranscriptClean as a PyPi library.

TranscriptClean v2.0.3

• pybedtools for split reads queries replace with pyranges.

TranscriptClean v2.0.2

• Added default value of False for primaryOnly and CanonOnly options. Python treated their default state as None, which was a problem.
• Added a check to fix_one_side_of_junction: in cases where the CIGAR string and the sequence of an edited read do not match, the function will now raise an error.

TranscriptClean v2.0.1

This is a patch that specifically affects the accessory script get_SJs_from_gtf.py.

• Updated syntax to Python 3.
• Removed requirement that GTF file contain the 'exon_number' field in the description as many GTF files lack this attribute.

TranscriptClean v2.0

• Added command-line option to limit output to canonical/annotated NCSJ transcripts. Also added tests covering this option.
• Added test files and dstruct.py file missing from v2.0-dev

From v2.0-dev

• Port code to Python 3.7.2
• Allow program to be run in multi-threaded fashion
• Prefilter splice junction and variant files in order to reduce memory requirements
• Restructured correction code to improve efficiency
• When reading from splice reference, allow noncanonical splice junctions that have short-read support but are not annotated
• Noncanonical splice junctions that are supported by the annotation are included as part of the non canonical total and non canonical corrected counts in the report and in the logs
• Add a more informative error message for fasta header issue
  Incorporate splice donor/acceptor status into reference junction matching
• Put tmp files into a designated directory and clean up after run if option is set
• Significant expansion of testing suite implemented in Pytest
• Minor changes to report script (for compatibility)
  TE log: Represent locations of indels using same convention as VCF format

TranscriptClean v2.0-dev

• Port code to Python 3.7.2
• Allow program to be run in multi-threaded fashion
• Prefilter splice junction and variant files in order to reduce memory requirements
• Restructured correction code to improve efficiency
• When reading from splice reference, allow noncanonical splice junctions that have short-read support but are not annotated
• Noncanonical splice junctions that are supported by the annotation are included as part of the non canonical total and non canonical corrected counts in the report and in the logs
• Add a more informative error message for fasta header issue
• Incorporate splice donor/acceptor status into reference junction matching
• Put tmp files into a designated directory and clean up after run if option is set
• Significant expansion of testing suite implemented in Pytest
• Minor changes to report script (for compatibility)
• TE log: Represent locations of indels using same convention as VCF format

TranscriptClean v1.0.8

Fixed bug reported by Matt Stone in which noncanonical splice junctions were corrected but changes to sequence and CIGAR did not make it to output because of a copied transcript object.

TranscriptClean v1.0.7

• Fixed deep copy syntax bug specifically affecting v1.0.6 that prevented non canonical splice junctions from being corrected. This was not a problem in other versions.
• Handles exception raised by certain non-canonical junctions in Oxford Nanopore data mapped using the GMAP aligner. PacBio data is typically not affected by this problem.

TranscriptClean v1.0.6

• Handles exception raised by certain non-canonical junctions in Oxford Nanopore data mapped using the GMAP aligner. PacBio data is typically not affected by this problem.
• Fixed bug affecting how unmapped reads were being written to the output SAM file