Releases: pybel/pybel
Releases · pybel/pybel
v0.15.5
v0.15.4
v0.15.3
v0.15.2
v0.15.1
v0.15.0
Added
- Support for homomultimers in ML triples export
- Graph dispatches allow quick access to PyBEL functionality (#449)
- Add "Streamable" BEL I/O (#451)
pybel.to_triples
allows direct export of ML-ready triples to a numpy ndarray.pybel.parse
function allows for quick parsing of BEL strings- Add JSON Schema for validating nodes and edges (#450) thanks @aman527
- Add BEL Repository functionality, previously located in https://github.com/pybel/bel-repository
- Pickling can be done with gzip to greatly reduce size both on files (d90578cb)
and on bytes (fcc99952)
Changed
- PyBEL now supports Python 3.6+ only.
pybel.to_tsv
renamed topybel.to_triples_file
- The citation entry in PyBEL JSON datastructure now uses
namespace
/identifier
/name
instead
ofdb
/db_id
. The corresponding constantspybel.constants.CITATION_DB
/pybel.constants.CITATION_DB_ID
/
pybel.constants.CITATION_DB_NAME
have been removed (#453). - The citation entry now relies on a subclass of
pybel.language.Entity
, which means empty strings are
no longer allowed. - The inconsistent usage of
subject
/source
as well asobject
/target
has been normalized everywhere.
This means the constantspybel.constants.SUBJECT
/pybel.constants.OBJECT
were removed and
new constantspybel.constants.SOURCE_MODIFIER
/pybel.constants.TARGET_MODIFIER
were added
(#453). - Remove the BEL default namespace. All usages get normalized to controlled vocabularies (mostly GO) automatically
(#455). - Improve namespace/annotation database insertion with Pandas (#454)
- Pickling now uses protocol 5 by default assisted by the
pickle5
backport on Python 3.6 (679dcab7) - Database now stores graphs in gzipped pickles (9ee9bf21)
- Unspecified molecular activities now use the top-level GO term (d56993e6)
- New annotation storage format using lists of annotations instead of dictionaries (#461)
- Citations now wrapped in dedicated data structure (#468)
Fixed
v0.14.10
Added
- Importer from EMMAA (#432)
- I/O for Amazon S3 (#431)
- Improve TSV exporter (d7d1287, 74d51c1, e108252, 6ffc1df)
- Add identifier-based entity remapper (ba8aa93)
- Add annotation grounding (#435, #443)
- Add HiPathia export examples (#422)
Changed
- Updated default BEL Commons instance from https://bel-commons.scai.fraunhofer.de to
https://bel-commons-dev.scai.fraunhofer.de - Add more namespaces for JGIF parsing to support CBN and BioDati import (9f74122 and
#435; thanks @djinnome) - Make Jupyter notebook export accessible from top level at
pybel.to_jupyter
(4d76faa)
Fixed
- Fix bug in display of nice labels in Jupyter notebook (775bdc3)
Removed
- Remove default service URL for BEL Commons exporter. The Fraunhofer instance was taken
down (a9a540f).
v0.14.9
v0.14.8
Changed
pybel.post_graphdati()
has been renamed topybel.to_biodati()
pybel.to_web()
has been renamed topybel.to_bel_commons()
pybel.from_web()
has been renamed topybel.from_bel_commons()
Added
- Content importers from GraphDati format with
pybel.from_graphdati()
andpybel.from_graphdati()
and enable usage of respective extensions*.bel.graphdati.json
and*.bel.graphdati.json.gz
with
pybel.load()
(#425) - Content importer from BioDati with
pybel.from_biodati()
(#425) - Add direct function for loading CBN files (so you don't have to load the JSON first) with
pybel.from_cbn_jgif_file()
- Added
pybel.grounding.ground()
function that uses the unlisted Python 3.7+ dependency, ``pyobo```
to ground/normalize entities in a given BEL graph. This also takes care of upgrading legacy namespace
names and mapping SCOMP/SFAM via FamPlex. (#426)
v0.14.7
Added
- Support for transcription factor relationships in TSV (machine learning) output