v0.1.0

First release of obnb, a Python toolkit for setting up benchmarking datasets using biomedical networks and gene annotations.

0.1.0-dev8

This is the last dev version for nleval. Next dev version will rename the project to obnb

What's Changed

• update install script (pytorch 2.0, pyg 2.3) by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/403
• feat: implement to_undirected_sparse_graph method for directed sparse graph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/404
• feat: use y_mask to enable considering selected negatives for training by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/408
• feat: orbital feature extraction extension module by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/416
• feat: pyt dataset object by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/421
• fix: pass log_level to internal processes by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/422
• add node_ids and task_ids to pyg data object by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/423

Full Changelog: krishnanlab/NetworkLearningEval@v0.1.0-dev7...v0.1.0-dev8

obnb/0.1.0-dev1

First OBNB dev release

v0.1.0-dev7

Summary

Feature

• Created the extension module nleval.ext to interface with other useful network learning tools. Currently support:
  • GRAPE
  • PecanPy
  • Scikit-Network

New

• Different channels from HumanNet (now HumanNet is downloaded from the original site, instead of through NDEx)
• Data release summary report will be generated, covering network and label stats

What's Changed

• fix: bring over node properties when restricting to branch by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/379
• fix: use factories for annotation and ontology data objects in annotated ontology by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/380
• fix: scanpy incompatibility with matplotlib 3.7 by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/381
• fix: add 'accession' scope to mygene converter to increase mapping coverage by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/385
• release: nledata-v0.1.0-dev5 (fix annotation data and go term info) by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/386
• fix: update network stats for testing by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/388
• feat: HumanNet download from original site; support different channels by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/391
• fix: add channel to config keys by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/392
• feat: implement ext module and add interfaces to grape and pecanpy by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/396
• fix grape imports and alignment by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/397
• feat: implement ext.sknetwork module by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/398
• make release data stats report by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/399
• release nledata-v0.1.0-dev6 data; update url and stats by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/400

Full Changelog: krishnanlab/NetworkLearningEval@v0.1.0-dev6...v0.1.0-dev7

v0.1.0-dev6

Summar

New

• Gene set collections
  • DisGeNET (and various sub-collections)
  • DISEASES (and various sub-collections)
  • HPO
• Network
  • ConsensusPathDB

What's Changed

• refactor annotated_ontology, break into annotation and ontology by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/370
• new: DISEASES gene set collection and DISEASESAnnotation by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/371
• new: HPO disease/trait gene set collection by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/372
• new: specific constructs of the DisGeNET and DISEASES gene set collections by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/373
• fix: annotated ontology data object redownload error by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/374
• fix: overloaded class docstring indentation error by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/375
• update: restrict mygene converter scopes by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/376
• new: implement the ConsensusPathDB network data object by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/377
• release: nledata-v0.1.0-dev4 by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/378

Full Changelog: krishnanlab/NetworkLearningEval@v0.1.0-dev5...v0.1.0-dev6

v0.1.0-dev5

Summary

New

• Networks
  • ProteomeHD
  • HuRI
  • HumanBaseTopGlobal
  • ComPPIHumanInt
  • OmniPath
  • HuMAP
  • SIGNOR
• Ontology
  • Replace DOID with MONDO

Updated

• CXExplorer update to improve cx file exploration
• DisGeNet -> DisGeNET

What's Changed

• update: add reduce option to eval_multi_ovr by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/306
• feat: random walk with restart model by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/307
• rename complete_node_attr to propagate_node_attr by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/309
• handles retries and 404 by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/313
• feat: download progress bar by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/314
• feat: to_df method for viewing lsc by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/315
• record avg time per epoch in gnn trainer by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/317
• add option to save training log file for trainers by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/318
• Create CODE_OF_CONDUCT.md by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/323
• setup api reference docs by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/324
• fix pyg install for docs by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/326
• create CONTRIBUTING.md by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/327
• rename DisGeNet -> DisGeNET by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/328
• allow changing data config options during data object initialization by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/330
• typo and format fixes by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/337
• update uuids for officially featured networks by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/341
• implement fields, node, and edge explorer methods by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/342
• add ProteomeHD and HuRI network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/346
• replace the human disease ontology with mondo ontology by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/347
• fix disgenet example: use mondo instead of doid by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/351
• use top 0.5% ProteomHD interactions as unweighted graph by default by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/352
• add HumanBaseTopGlobal network, implement new BaseURLSparseGraph object by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/354
• implement ComPPIHumanInt network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/355
• implement OmniPath network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/356
• add HuMAP network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/357
• add SIGNOR network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/358
• update actions versions by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/359

Full Changelog: krishnanlab/NetworkLearningEval@v0.1.0-dev4...v0.1.0-dev5

v0.1.0-dev4

What's Changed

• Bump numpy from 1.22.4 to 1.23.2 by @dependabot in https://github.com/krishnanlab/NetworkLearningEval/pull/279
• fix: nonred filter by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/293
• fix: obo read is_a and part_of relationships by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/294
• update release data script by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/295
• bump data version nledata-v0.1.0-dev1 -> nledata-v0.1.0-dev2 by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/296
• update: do not use edge weights in SimpleGNNTrainer by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/297
• feat: metrics reduction by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/298
• update: pass label and splitter to construct Dataset by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/299
• feat: eval_multi_ovr method for sl and lp trainers by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/300
• build(deps): bump numpy from 1.23.2 to 1.23.3 by @dependabot in https://github.com/krishnanlab/NetworkLearningEval/pull/301
• refactor: train and eval_multi_ovr for standard trainers by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/302
• feat: dataset_constructors by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/303

Full Changelog: krishnanlab/NetworkLearningEval@v0.1.0-dev3...v0.1.0-dev4

v0.1.0-dev3

bump version: 0.1.0-dev2 -> 0.1.0-dev3

0.1.0 beta

Summary of some main changes

• New features
  • Archive data version, allowing full reproducibility
  • Non-redundant label set collection filtering
  • Save data processing config file
  • Filter gene annotation data sources
• Several refactoring
  • GenePropertyConverter
  • Dataset
  • Feature
• Improved graph API
  • remove_edge
  • add_nodes
  • add_node
  • get_neighbors

What's Changed

• Fix data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/155
• Clone NDEx networks and timout DisGeNet test by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/158
• Induced subgraph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/162
• Data nosave by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/164
• Connected component by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/165
• Fast sparse subgraph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/167
• CX network data preprocessing with connected comp by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/169
• Fix pbar by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/173
• Log dataset downloading/processing by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/174
• Cache transformed data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/177
• Add PCNet v1.3 network by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/179
• NDEx CX stream data explorer by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/180
• Only preserve human genes in network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/182
• Remove genes with ambiguous Entrez mapping by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/183
• Cache mygene conversion by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/184
• Schedule example execution test instead of PR/push by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/185
• Test converter by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/187
• Sort imports via isort, line-length 80 -> 88 by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/190
• Fix yaml format by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/191
• Fix absolute imports by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/205
• Implement non-redundant labelset filters by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/216
• Implement experimental dataset object AlevinFry by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/220
• Implement archival data download option by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/222
• update: get_nonred_label_ids use sort tuples by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/229
• Check data transform configuration by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/230
• Refactor pre-transform by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/231
• Improve transformed data caching and logging by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/232
• Data object config file by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/233
• Set up bump2version by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/236
• Implement config checker by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/237
• set up node id converter using preset name by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/238
• update data config format; fix network stats by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/239
• Release data nledata-v0.1.0-dev and test data on latest release by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/243
• update: test examples by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/244
• update: graph api add_node(s) methods by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/245
• feat: get_neighbors graph api method by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/246
• feat: remove_edge methods for sparse graph objects by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/247
• refactor config by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/248
• refactor: exceptions by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/249
• refactor: feature by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/250
• fix: patch graphgym by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/251
• Refactor Dataset out from model_trainer by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/253
• merge y and masks into Dataset by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/255
• use archived data in examples by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/260
• feat: GenePropertyConverter by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/261
• refactor: split and property_converter by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/265
• update: download versioned data cache if not it is not avaialbe yet by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/269
• update: DisGeNet curated annotation source by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/272
• update: GO curated annotation source by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/273
• update readme with updated api by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/275
• Update data release by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/277

Full Changelog: krishnanlab/NetworkLearningEval@v0.0.3...v0.1.0-dev1

Data objects, GraphGym trainer, OntologyGraph and more

What's new

• MultiFeatureVec for sample classification and hyperparameter tuning uses
• OntologyGraph and DirectedSparseGraph that keeps tract of reversed edge direction on a directed graph for reversed propagation
• Data objects NLEval.data, including several network data and two annotated_ontology data.
• GraphGymTrainer
• Logger and progress bar

What's Changed

• Update MultiFeatureVec methods by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/91
• Multi trainer, add hyperparam tuning example usage by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/92
• Docstring by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/93
• Featurevec align by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/94
• MultiFeatureVec prediction tasks such as sample classification by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/96
• Load graph from CX stream file by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/97
• Load edge weights from edge attributes for CX by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/98
• Network data retrieval via NDEX by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/99
• Edge reduction by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/102
• Network data preprocessing for efficient loading by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/103
• Gene ID converter by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/105
• Test npz readwrite and network data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/106
• Refactor examples by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/107
• Refactor types by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/109
• Ontology graph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/110
• Read obo and construct ontology graph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/111
• Download and process the DisGeNet data by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/114
• Filter progress bar by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/115
• Refactor Jaccard and Overlap pairwise score based labelset filtering by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/118
• GeneOntology data object by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/120
• Self loops in SparseGraph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/124
• Data reprocess/redownload option by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/127
• Refactor graph and add conversion factory between SparseGraph and DenseGraph by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/129
• Create GraphGymTrainer by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/130
• Example execution Github Action by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/136
• Rewind GraphGym model to optimal state by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/137
• Logging by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/142
• Trainer logger by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/143
• Filter logs by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/144
• Refactor typing (centralize to NLEval.typing) by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/146
• Type EdgeData; require explicit setting of property_name by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/148
• Update logger by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/149
• Refactor data objects by @RemyLau in https://github.com/krishnanlab/NetworkLearningEval/pull/150

Full Changelog: krishnanlab/NetworkLearningEval@v0.0.2...v0.0.3