Release v0.5.0

Summary:

Major

File reading changed to AlphaRaw

• This allows SciEx support. FF is not yet optimized for SciEX yet
• Adjusted memory limits for raw conversion, raw conversion for Bruker is now faster
• Bruker legacy files are supported with the one-click installer but not for the Python version anymore. Install python-lzf or install with the legacy requirements,
• Working docker files

Minor

• Option to change the top_n peptides
• Option to select the initial score for ML scoring
• Fixed a bug in peptide generation that would not consider the last AA
• Added a new protease protease_dict["trypsin/p"] = "[KR]"
• Small UI improvements

Performance:

• Overall, Bruker runs should be 25% faster. We will expect further speed increases once the FF is exchanged
• IDs dropped ~5%, but FDR estimates are well below 1% now

Full Changelog: v0.4.9...v0.5.0

Release v0.4.9

What's Changed

• Upgraded to NBDEV 2.0
• More documentation
• Better display of AlphaPept version number

Full Changelog: v0.4.8...v0.4.9

Release v0.4.8

What's Changed

• Multiple small optimizations
• Plot for precursor calibration, general optimization of the calibration
• New generic score to replace the x_tandem
• Memory optimizations
• Changed the default prec_tol to 20, default_mass_tol to 50
• #488: fix error when scoring on machines with more than 60 cpus by @mschwoer in #489

Full Changelog: v0.4.7...v0.4.8

Release v0.4.7

What's Changed

0.4.7

• Updated multiple packages & documentation
• Print traceback when there is an error running AlphaPept
• Updated modifications.tsv, Button to reload modifications via GUI
• Changed slider input widget to a number input
• Fixed links in GUI when downloading FASTA
• Fixed GUI Errors #460
• Fixed bug when launching file watcher
• Option to use relative file paths for the size check @hugokitano
• New functionality to filter fragments by local rank. This should improve performance for noisy datasets.
• Cleaned up print output/deprecation & performance warnings
• Macbook crash fix by @swillems in #465
• FEAT: create headless option by @swillems in #464 #449
• Develop by @straussmaximilian in #478

New Contributors

• @k-Vartika made their first contribution in #450
• @elena-krismer made their first contribution in #454

Full Changelog: v0.4.6...v0.4.7

Release v0.4.6

What's Changed

• Link to Renku @cmdoret #433
• Optimized quantification (proper integration of signal on MS1 level), now uses apex of isotope envelope
• Caching of data frames for matching @mschwoer #438
• Minor code reorganization
• Bugfixes in interface @hugokitano
• Initial estimates for delayed normalization
• Bugfix for Threadripper Systems #434, #264
• Updated docs
• More flexibility for custom modifications, see constants-tab in GUI
  

Full Changelog: v0.4.5...v0.4.6

Release v0.4.5

What's Changed

Revision of how fractions are handled

• This should improve performance on fractionated samples #424 #389
• Scoring was adapted. For fractionated samples each raw file is scored individually.
• In the GUI there are now two additional columns: Sample_group and Fraction: All files with the same Sample_group will be quanted together. Fraction is meant to be enumerated for fractionated samples and is used for the delayed normalization step.

Acceleration

• @hugokitano made their first contribution in #412
• This is a speedup for the delayed normalization. Thanks!

Misc

• The version performance is updated and shows results from version 4.
• Automated test now include fractionated samples and matching.
• Pip installation on Windows should be more robust (fixed tables dependency)

Package Updates

Several Python packages have been updated:

• Bump addressable from 2.7.0 to 2.8.0 in /docs by @dependabot in #269
• Bump matplotlib from 3.3.3 to 3.5.1 by @dependabot in #376
• Bump scikit-learn from 0.24.2 to 1.0.2 by @dependabot in #380
• Bump h5py from 3.4.0 to 3.6.0 by @dependabot in #384
• Bump pyinstaller from 4.2 to 4.10 by @dependabot in #415
• Bump pyteomics from 4.4.2 to 4.5.3 by @dependabot in #416
• Bump pandas from 1.3.4 to 1.4.2 by @dependabot in #411
• Bump sqlalchemy from 1.4.23 to 1.4.35 by @dependabot in #413
• Bump streamlit-aggrid from 0.2.2.post4 to 0.2.3.post2 by @dependabot in #417
• Bump pyyaml from 5.4.1 to 6.0 by @dependabot in #418
• Bump tqdm from 4.56.0 to 4.64.0 by @dependabot in #419
• Bump scipy from 1.7.1 to 1.8.0 by @dependabot in #421
• Bump networkx from 2.6.3 to 2.8 by @dependabot in #422
• Bump pyinstaller from 4.10 to 5.0 by @dependabot in #420

Full Changelog: 0.4.4...v0.4.5

Release 0.4.4

This summary includes the changes from the previous .4.xx:

0.4.4

• New maximum iterations for LFQ: This should avoid the "not converged"-messages when doing LFQ
• Auto-adjustment of settings for non-specific search: This should avoid memory errors when using non-specific as protease

0.4.3

• Bugfix for LFQ

0.4.2

FEAT: Better logging for LFQ. This will now log if a solver does not converge.
FIX: Correct setting of cores when searching w/o saving database.
DOCUMENT: Updated DOCS

0.4.1

FIX: Fixed a bug in LFQ that made LFQ take very long.
0.4.0

This version contains a lot of variable renaming to be more consistent with other Alpha*-packages.
This will probably lead to incompatibility with previous history modes.

Optimized recalibration to be more memory efficient
Fixed a bug where peptide_fdr was not accurately saved. This should affect reported protein / peptide ids.
Included additional export of identifications, e.g. the best sequence for each recorded MS2-spectrum
Usability improvements when starting AlphaPept multiple times
Revised History Mode

Release v0.4.2

This also includes the changes from the previous .4.xx:

0.4.2

• FEAT: Better logging for LFQ. This will now log if a solver does not converge.
• FIX: Correct setting of cores when searching w/o saving database.
• DOCUMENT: Updated DOCS

0.4.1

• FIX: Fixed a bug in LFQ that made LFQ take very long.

0.4.0

This version contains a lot of variable renaming to be more consistent with other Alpha*-packages.
This will probably lead to incompatibility with previous history modes.

• Optimized recalibration to be more memory efficient
• Fixed a bug where peptide_fdr was not accurately saved. This should affect reported protein / peptide ids.
• Included additional export of identifications, e.g. the best sequence for each recorded MS2-spectrum
• Usability improvements when starting AlphaPept multiple times
• Revised History Mode

Release v0.4.1

Description will follow shortly

Release v0.4.0

Description and Update will follow shortly