v1.0.3

What's Changed

• build: bump version to v1.0.3 by @BioGeek in #122

Full Changelog: v1.0.2...v1.0.3

v1.0.2

What's Changed

• ci: normalize package name to canonical name by @BioGeek in #120

Full Changelog: v1.0.1...v1.0.2

v1.0.1

What's Changed

• ci: change project name to winnow_fdr to match PyPI name by @BioGeek in #118

Full Changelog: v1.0.0...v1.0.1

v1.0.0

What's Changed

• fix: filter out sequences longer than 30 in first and second beam prediction by @JemmaLDaniel in #54
• feat: add z-score over beam results to beam features by @JemmaLDaniel in #52
• feat: add fdr metrics into metadata table for bayes and database-grounded classes by @JemmaLDaniel in #48
• fix: filter out first beam sequences longer than 30 amino acids by @JemmaLDaniel in #55
• Feat create a train and predict script by @JemmaLDaniel in #51
• fix: fix database-ground confidence-cutoff by @JemmaLDaniel in #64
• fix: drop top predictions when computing fdr threshold by @JemmaLDaniel in #63
• Feat: change from logistic regression to MLP classifier by @JemmaLDaniel in #68
• Fix confidence cutoff to consistently select leftmost root by @JemmaLDaniel in #66
• chore: add license by @JemmaLDaniel in #83
• Fix add graceful error handling by @JemmaLDaniel in #82
• Feat dockerise for AIchor by @JemmaLDaniel in #76
• feat: add unit tests by @JemmaLDaniel in #81
• CI: run tests on merge into main by @JemmaLDaniel in #100
• fix: set irt predictor seed to 42 by default by @JemmaLDaniel in #93
• chore: add external dataset modifications to residue masses dict by @JemmaLDaniel in #92
• fix: only save an irt predictor model when available by @JemmaLDaniel in #94
• feat: add q-value support by @JemmaLDaniel in #95
• refactor: enforce same FDR calculation methods between database-grounded and non-parametric methods by @JemmaLDaniel in #103
• Feat: use non-parametric FDR estimation by @JemmaLDaniel in #69
• feat: add casanovo dataset loader by @JemmaLDaniel in #75
• feat: add early stopping to the calibrator by @JemmaLDaniel in #104
• refactor: implement DatasetLoader interface for data loading by @JemmaLDaniel in #73
• feat: create winnow example notebook by @JemmaLDaniel in #84
• 96 add mkdocs by @JemmaLDaniel in #97
• fix: make scripts folder a package by @BioGeek in #109
• refactor: explicitly specify packages for setuptools by @BioGeek in #111
• docs: fix workflow image URL in README by @BioGeek in #113
• publish the winnow_fdr package to pypi by @BioGeek in #115
• docs: add link to preprint in notebook and README by @BioGeek in #116

New Contributors

• @BioGeek made their first contribution in #109

Full Changelog: v1.0.0a...v1.0.0

v1.0.0a

What's Changed

• Port code from IN repo by @amabinsta in #1
• 11 implement to csv method to calibrationdataset class by @JemmaLDaniel in #12
• add bug and feat issue templates by @JemmaLDaniel in #16
• add pre-commit config by @JemmaLDaniel in #15
• fix: correct type hint for compute_ion_identifications by @JemmaLDaniel in #23
• resolve dependencies by @JemmaLDaniel in #17
• Implement alternative fdr scoring metrics by @JemmaLDaniel in #13
• Fix confidence cutoff bug by @amabinsta in #26
• fix: fix indexing errors for chimeric and beam features by @JemmaLDaniel in #21
• create winnow cli by @amabinsta in #35
• fix: swap numpyro beta parameters to correct ones by @JemmaLDaniel in #44
• docs: update README with project overview, installation instructions,… by @amabinsta in #39
• fix: prevent division by zero in compute_fdr by @JemmaLDaniel in #38
• feat: add a to_parquet method for metadata saving by @JemmaLDaniel in #37
• feat: add GitHub Actions workflows by @amabinsta in #40
• feat: improve issue templates by @amabinsta in #42
• adapt winnow to work with new instanovo by @JemmaLDaniel in #33

New Contributors

• @amabinsta made their first contribution in #1
• @JemmaLDaniel made their first contribution in #12

Full Changelog: https://github.com/instadeepai/winnow/commits/v1.0.0a