v0.5.16-rc1

Full Changelog: v0.5.15...v0.5.16-rc1

v0.5.15

Pydantic v1 and v2 compatibility

This release of OAK should be compatible with both v1 and v2 of pydantic

Thanks to @hrshdhgd and @cthoyt for work on downstream libraries that made this possible!

What's Changed

• Making pydantic v1 vs v2 neutral by @cmungall in #632

Full Changelog: v0.5.14...v0.5.15

v0.5.14

This release is explicitly only compatible with pydantic<2. Previously it was possible to install both pydantic v2 and OAK yet run into compatibility issues (when importing bioregistry it auto-creates objects with Pydantic v1 assumptions).

Note we are working hard to make a new release that will explicitly work with both v1 and v2.

This release is also pinned to semsimian 0.1.18

What's Changed

• Pinning pydantic to be <2 until we are sure the full stack works. See #628 Bumping semsimian to 0.1.18 by @cmungall in #630

Full Changelog: v0.5.13...v0.5.14

v0.5.13

Highlights

Fill table now allows for obsoletion rewiring

PR: #625

This fixes #625

given a table of single-cell data sc.tsv:

dataset	term
sc-Martinotti	CL:4023067
sc-Martinotti	CL:4023027

Running

runoak -i sqlite:obo:cl fill-table --relation "{primary_key: term, dependent_column: replacement_term, relation: IAO:0100001}" --allow-missing tests/input/cl-obs.tsv sc.tsv

Will produce:

dataset	term	replacement_term
sc-Martinotti	CL:4023067	PCL:1000004
sc-Martinotti	CL:4023027

Inclusion of ABox some-values-from assertions

PR: #624

Previously axioms like the anonymous Type assertion here we not included in the graph projection:

Now they are:

Logical Definition Analysis

• #610

What's Changed

• Make ensure_strict_prefixes default to False in lexmatch by @hrshdhgd in #601
• Improve descriptor processing for Gilda implementation by @cthoyt in #603
• Enable NER with Gilda by @cthoyt in #607
• Tools to analyze logical definitions by @cmungall in #610
• Fix broken link on README.md by @lubianat in #599
• Projecting anonymous type expressions to graph. Fixes #623 by @cmungall in #624
• Adding arbitrary properties to fill-table. Fixes #625 by @cmungall in #626

New Contributors

• @lubianat made their first contribution in #599

Full Changelog: v0.5.12...v0.5.13

v0.5.13-rc4

Full Changelog: v0.5.13-rc3...v0.5.13-rc4

v0.5.13-rc3

Full Changelog: v0.5.13-rc2...v0.5.13-rc3

v0.5.13-rc2

Full Changelog: v0.5.13-rc1...v0.5.13-rc2

v0.5.13-rc1

What's Changed

• Make ensure_strict_prefixes default to False in lexmatch by @hrshdhgd in #601
• Improve descriptor processing for Gilda implementation by @cthoyt in #603
• Enable NER with Gilda by @cthoyt in #607

Full Changelog: v0.5.12...v0.5.13-rc1

v0.5.12

What's Changed

• Semsimian handles thresholds for resnik and jaccard by @hrshdhgd in #597

Full Changelog: V0.5.11...v0.5.12

V0.5.11

What's Changed

• minor semsim optimizations by @cmungall in #582
• Use prod api url for monarch_implementation by @kevinschaper in #591
• Add benchmark for Phenio HP vs MP by @caufieldjh in #578
• Implement all_by_all_pairwise_similarity in semsimian implementation by @justaddcoffee in #579
• Allow export of dot files and cx files from viz by @cmungall in #594
• Workaround for ontoportal prefix issue 592 by @cmungall in #593
• Applying thresholds to avoid expensive object production. by @cmungall in #595

New Contributors

• @kevinschaper made their first contribution in #591

Full Changelog: v0.5.10...V0.5.11