Skip to content

Version 1.5.0

Latest
Latest
Compare
Choose a tag to compare
@douweschulte douweschulte released this 19 Feb 16:19
· 4 commits to master since this release
v1.5.0
d17c6b6
This commit was signed with the committer’s verified signature.
douweschulte Douwe Schulte
GPG key ID: 30831308C656B7D7
Learn about vigilant mode.

Update paper (since 1.0) (preprint)

Updates

  • Added loading of sequences from mmCIF files, for ModelAngelo (#207, #217)
  • Added loading of sequences from Peaks 11 and Peaks Ab
  • Added loading of sequences from Casanovo, pNovo, and MaxNovo (#192)
  • Added loading of spectra from Casanovo, pNovo, and MaxNovo (#195)
  • Added I/L disambiguation based on satellite ions (#193, #216, #218)
  • Added more customization options for peptide fragmentation
  • Added hash for all used files in the HTML report (#229)
  • Added the option to call parts of a batchfile from another file (include!(<path>)) see the examples (#111, #209)
  • Added depth of coverage information to the Fasta export
  • Updated the main overview header (#212)
  • Updated the command line interface (CLI) to be more friendly, Note: calling stitch has changed to stitch run <path>
  • Updated the batchfile parsing to provide more helpful error messages
  • Updated the segment overview tree, always use Blosum62, ability to not generate it (BuildTree: False) (#213)
  • Updated the exact scoring of reads to handle longer alignments better, as well as locally enforcing unique placement (#215)
  • Updated segment joining to use the same alignment as template matching (#128)
  • Deprecated the use of booleans with EnforceUnique, for now will keep working with a warning but support will be dropped at some point.
  • Fixed small remaining intensity [1-2] scaling for Novor reads to scale [0-1]
  • Fixed issues with the generation of the consensus sequences in relation to the mass alignment and I/L disambiguation (#220)
  • Fixed leading insertions in reads placement displayed in a different colour, for more clarity in the alignment
  • Fixed bugs (#193, #222, #230, #242, #243, #244)
  • Performance improvements
  • Many more small fixes and improvements

Note

  • The benchmarks show a mostly very slightly positive result, with the caveat that I/L disambiguation is not automatically benchmarked yet because raw files are so big.
  • Performance has been improved with a couple of tiny steps. The I/L disambiguation though adds extra work and so results in longer runs when used, but this scaling is linear with the number of input reads. The move to net7.0 also gives a bit of performance improvement.

Breaking changes

  • The CLI has changed: stitch <batchfile> has been changed to stitch run <batchfile> to better group the applicable arguments for all subcommands. You can use --help to get general help or <subcommand> --help to get help on that subcommand.
  • The CSV export has been altered by adding more columns with I/L disambiguation specific information. The order of the other columns has not been changed.

Known issues

  • The segment annotation can be off by a couple of aminoacids, or it can miss segments after CDR3.

See the rest of the changelog in CHANGELOG.md.

Assets 6
Loading