Changelogs » Deepchem

PyUp Safety actively tracks 323,951 Python packages for vulnerabilities and notifies you when to upgrade.

Deepchem

2.5.0

See full release notes at https://forum.deepchem.io/t/deepchem-2-5-0-release/439

2.4.0

Read the full release notes at https://forum.deepchem.io/t/deepchem-2-4-0-release-notes/340

2.3.0

This release of DeepChem swaps from our home-grown `TensorGraph` framework to using Keras as the foundation of our models. This swap leaves us well prepared for the jump to Tensorflow 2.0 which will happen in our next major release. This version also bumps the TensorFlow version to 1.14. This release also includes a number of improvements to MoleculeNet and our transfer learning infrastructure
  
  Remove uses of deprecated APIs 1550
  Added attr-slow for the AtomicConvFeaturizer test 1552
  Upgrade to TensorFlow 1.13.1 1553
  fix bug of load_pdbbind() and add new features 1561
  Replaced Saver with Checkpoint 1566
  Replaced uses of deprecated layers 1567
  Convert TensorGraph layers to Keras layers 1578
  Create KerasModel 1583
  Update dependencies for DeepChem 2.2 1584
  Converted multitask models to KerasModel 1587
  Remove contagious logger setup 1591
  Converted graph models to KerasModel 1594
  Construct dataset first time, even with reload set to True 1595
  Loading thermosol and hppb datasets 1596
  simple install one-liner 1602
  Converted more models to Keras 1615
  Smiles Based featurizers for ChemNet 1618
  Converted progressive multitask models to KerasModel 1620
  Swapping Split-Transform order 1621
  Added ChemNet models with tests 1623
  Swap Split-Transform order - II 1624
  Converted GAN to KerasModel 1625
  Converted reinforcement learning classes to Keras 1635
  Created new MAML API 1636
  SmilesToImage featurizer for Tox21, Sampl, HIV datasets 1637
  ChemNet Fixes and Additions 1638
  First version of pretrained loading 1643
  Upgrade to TF 1.14 (Optional) 1645
  Custom directories and SmilesToImage for MolNet 1649
  ChemNet Fixes 1651
  Created ValidationCallback 1652
  Moved to Python 3.5 and 3.7 for Travis 1658
  Stratified splitters, and minor changes for MolNet 1660
  Updated installation instructions 1661
  Workaround for bug in TF 1.14 1662
  Reorganized models directory 1664
  Move test cases out of tensorgraph module 1666
  Fixed broken and out of date examples 1671
  Updated version number to 2.3.0 1672
  Update README.md 1682
  DiskDataset.move() would not overwrite an existing dataset 1683

2.2.0

DeepChem 2.2 takes large steps towards making DeepChem a general purpose deep learning library for life science applications. Major improvements have been made to support for deep learning on protein structures, and significant support for image-based dataset and model handling has been added. In addition, tooling for interpreting deep models has been improved. A number of improvements to existing models have been added as well, including adding estimator support for a number of model classes. Many bugfixes and small improvements made it in as well. DeepChem 2.2 now depends on TensorFlow 1.12.
  
  
  PDBBind and Protein Structure Improvements
  1366, 1383, 1411, 1413, 1476 Atomic Convolution Improvements
  1503, 1430, 1432 PDBBind bugfixes
  1497 Using binding pockets to load PDBBind
  1498 DeepMHC for protein peptide binding
  1369, 1360, 1372, 1397 Featurziation Improvements
  1498 DeepMHC for protein peptide binding
  
  Image Handling Improvements
  1516 Image Transformation improvements
  1324 Cell counting dataset added. `ImageLoader` added
  1414 Diabetic Retinopathy example model
  1439 `ImageDataset` class
  
  Dataset and Splitter Additions, Improvements and fixes
  1507, 1540, 1406 Bugfixes
  1514 Handling verbose=False when transforming data
  1499 Butina splitter improvement
  1347 Adds USPTO dataset.
  1348 BBBC002 dataset addition
  1339 Split datasets on ID
  1416 Molnet loaders for UV/Kinase/Factors datasets
  1327 BBBC001 dataset addition
  1447 SDFLoader improvements
  1425 Binary classification metric improvements
  
  Model and Layer Additions and Improvemetns
  1500 Seq2seq model improvements
  1513 Clean up symmetry functions
  1488 New graph convolution
  1365 Average pooling for conv-nets
  1370 ResNet50 improvements
  1450 Layer output shapes
  1452 Pad batch improvements
  1453 Example distributed multitask classifier
  1335 GraphConv improvements
  1343 Making it easy to pull out neural fingerprints
  1433 TensorGraph get layer weights
  1325 UNet model changes
  1334 First Resnet50 build
  1473, 1142 TextCNN make_estimator support
  1475 DTNN make_estimator support
  1495 ANIRegression, BPSymmetryFunction make_estimator
  
  Better Interpretability
  1393 Saliency Mapping
  1445 Saliency maps for diabetic retinopathy
  
  Tests, Docs, Housekeeping
  1527, 1457 Readme cleanup
  1548 Version bump

2.1.0

This release of DeepChem includes major upgrades to the TensorGraph framework by enabling support for TensorFlow Eager and Uncertainty estimation. TensorGraph layers can now be used in Eager mode to build dynamic models. In addition, a number of TensorGraph models, notably regression models, have been upgraded to allow for uncertainty estimates in model predictions. In addition, this release features a host of other improvements and bugfixes to `TensorGraph`, including support for saving submodels and significant refactoring to simplify underlying code. We've also simplified the names of `TensorGraph` models: `TextCNNTensorGraph` is now `TextCNN` with similar changes for other models. Old names are still supported, but deprecated and will be removed in DeepChem 3.0. We've also added a number of new IPython notebook tutorials explaining new DeepChem features.
  
  This release also lays some groundwork for upgraded support for and image based datasets that will be coming in future releases.
  
  
  - Eager Mode and Uncertainty Estimation
  - Eager Mode support 1191 1176
  - Uncertainty estimation support 1226 1218
  - TensorGraph Improvements and Fixes
  - TensorBoard improvements 1295 1261
  - TensorGraph refactoring 1286 1259 1196
  - TensorGraph Submodel saving support 1264
  - TensorGraph model improvements 1274 1229 1195 1181 1178 1174 1167
  - Improved TensorGraph Estimator support 1153
  - PowerSign optimizer 1233
  - Hinge loss improvements/fixes 1242 1189
  - Rename `MultiTask` to `Multitask` 1249
  - New TensorGraph Models
  - TensorGraph UNet Implementation 1272
  - TensorGraph Progressive Multitask Classifier 1151
  - Example/Tutorial Improvements
  - Example/MoleculeNet Fixes  1310 1308 1307 1305 1298 1296 1251 1227 1215 1183 1141 1131
  - New IPython Notebook Tutorials
  - `tf.data` and `tf.estimator` Notebook 1136
  - Lime Notebook (model agnostic method explanation) 1202
  - Synthetic feasibility scoring notebook 1182
  - Initial support for Image and Genomics based models
  - New Image Transformations API 1300
  - Improved support for genomic applications 1313 1278
  - Builds, tests, and bugfixes
  - Build improvements 1291 1290 1276 1221 1201 1165 1134 1132 1320
  - Miscellaneous Bugfixes/Cleanup 1228 1164 1145 1135
  - Python 3 Fixes 1235
  - Fix NumpyDataset reshaping 1260
  - Featurization Improvements
  - Hashable Featurizers 1267
  - Atom based features support for graph-convs 1156

2.0.0

This major version release finishes consolidating the DeepChem codebase around our TensorGraph API for constructing complex models in DeepChem. We've made a variety of improvements to TensorGraph's saving/loading features and added a number of new tutorials improving our documentation of TensorGraph. We've also removed a number of older deprecated submodules and models in favor of the new, standardized TensorGraph implementations.
  
  In addition, we've implemented a number of new deep models and algorithms, including DRAGONNs, Molecular Autoencoders, MIX+GANs, continuous space A3C, MCTS for RL, Mol2Vec and more. We've also continued improving our core graph convolutional implementations.
  
  We've also made a variety of documentation, build, and website improvements and fixes.
  
  Our thanks to our contributors for all the hard work!
  
  - TensorGraph Conversion and Upgrades
  - 925 Sequential API for Model Construction
  - 949 Cleanup of Examples plus bugfixes
  - 952 Using Queues for prediction
  - 954  GraphConvTensorGraph Classification with Multiple Tasks
  - 965 Can restore from any checkpoint
  - 966 Simplified names of TensorGraph graph conv models
  - 970 Added TensorGraph LSTM Layer
  - 972 Add Model Configuration params to GraphConvTensorGraph
  - 975 Swap examples to use new TensorGraph models
  - 976 Remove deprecated TensorFlow classes
  - 989 Remove `tf_new_models` old submodule
  - 992 Remove the deprecated old fully connected models
  - 1007 Add ability to change loss functions after reload
  - 1023 Add TensorGraph Cast layer
  - 1024 Can move saved TensorGraph models on disk
  - 1054 Implement IRV TensorGraph Model
  - 1082 Move `tensorflow_models` and `autoencoder_models` into contrib
  - 1083 Add one-shot code back into contrib
  - 1085 Saving and loading Weave models
  - 1086 Implementation of robust models in TensorGraph
  - 1090 Move `dc.nn` to contrib
  - Reinforcement Learning Upgrades
  - 931 Monte Carlo Tree Search for RL
  - 1022 A3C supporting continuous action spaces.
  - Graph Convolution Improvements:
  - 1033 Adding master atoms to graph convolutions
  - 1080 Adding chirality to Atom and Bond features
  - 1081 Pinning Graph Gather to CPU for TF Bug
  - 1105 Graph Normalization
  - MoleculeNet Improvements
  - 933 Run MoleculeNet on Jenkins
  - 958 Building multi-assay datasets from PubChem based on genes
  - 996 PCBA dataset generation based on a single gene
  - 1032 Simplify Tox21 loading
  - 1042 Update MoleculeNet to latest models
  - 1049 Fixes for MoleculeNet update
  - New models added to DeepChem
  - 939 Mol2Vec implementation added to contrib
  - Add DRAGONN models
  - 979 Adding DRAGONN example to contrib
  - 1003 Removing some commented out DRAGONN code
  - 1008 Adding DNA Simulation code
  - 1020 Implementing FASTALoader
  - 981 MIX+GAN implemented
  - 1026 Molecular Autoencoder Implementation in TensorGraph
  - dc.data improvements
  - 930 Complete Shuffle Disk Dataset
  - 1031 Bugfix for merging DiskDatasets
  - 1034 Enabling merging of NumpyDataset
  - 1091 Adding `dataset.make_iterator` to create `tf.data.Iterator` instances from Dataset objects
  - Improvements to featurization and data splits
  - 1005 Enable choice of featurizer to be searched as a hyperparameter
  - 1009 Add Maxmin splitter
  - DeepChem tutorial additions and improvements
  - 940 Update datasets for protein-ligand complex tutorial
  - 961 Fix BACE tutorial
  - 998 Fix GraphConv tutorial
  - 1051 New TensorBoard usage tutorial
  - 1104 New splitter tutorial
  - 1072 DeepChem Organizational Structure and Governance
  - Documentation improvements
  - 957 Bump to 1.3.1
  - 997 Docstring for `dc.utils.load_from_disk`
  - 1018 Add README link
  - 1050 Improve README
  - 1053 README fixes
  - 1067 Docs fix to keep numpy docstrings rendering
  - 1068 Update SAMPL example docstrings to read correctly
  - 1077 Fix README links
  - 1107 Better installation from source examples
  - 1117 Fixes to prevent test modules from being generated in docs
  - 1123 Convert some leftover GPL license tags to MIT
  - 1124 Version bump for 2.0.0
  - Build Improvements
  - 928 Bump conda, docker to 1.3.1 version
  - 1013 Add manifest to include data files
  - 1028 Install simdna from conda-forge
  - 1056 Turn off PyPi uploading to deal with bugs
  - 1059 TensorFlow 1.5 Upgrade
  - Website improvements
  - 932 Bump Website version
  - 1057 Number of website fixes
  - Bugfixes, Tests, and Miscellaneous Improvements
  - 934 Upgrade yapf version to 0.19
  - 945 Fix import bug
  - 948 Write values not tensors to TensorBoard
  - 964 Fix infinite loop caused by shards of size 0
  - 973 Fix featurizer name error
  - 982 Fix import errors from ICU package
  - 987 Speed up tests
  - 1001 Upgrade yapf version to 0.20
  - 1061 Fix gaussian process optimization bug
  - 1063 Install pbr package properly
  - 1088 Fix import bugs
  - 1113 Using `logger` instead of print

1.12

1462 Typo fix predict_proba
  1385, 1418 Build Fixes
  1423 Yapf updates
  1408 indentation cleanup
  1344 Python 3.6 updates
  1330: Docs updates
  1337 Large screens tutorial
  1338 Colab notebook version
  1437 Python3 fixes
  1454 Make RDKit a soft requirement
  1455 Make simdna a soft requirement
  1456, 1484  Make six a soft requirement
  1458 Add tutorial section
  1420 genomics code grouping into single file
  1535, 1485, 1487, 1371, 1421, 1479, 1480 Test Improvements and Fixes

1.3.1

This minor release swaps the backend metadata storage for `DiskDataset` to hd5 from pickle files. This change was necessary since pickle is a brittle file format. (A recurring issue was that py2 pickles and py3 pickles were not compatible). This minor release introduces a transition implementation: loading any `DiskDataset` on disk with the `dc.data.DiskDataset` implementation in this release will automatically swap from pickle to hd5 format  on disk with no extra input needed from users.
  
  The next DeepChem release is going to be a 2.0 release which will consolidate on the hd5 format, so we encourage users to upgrade to 1.3.1 and transition any data stored on disk in preparation.
  
  Thanks as always to our core developers and special thanks to our first-time contributors!
  
  Detailed Changes:
  - Pickle to hd5 transition for `DiskDataset` (899)
  - TensorGraph Improvements (895, 901, 917)
  - Dataset fixes (914, 916)
  - MoleculeNet Jenkins build (910)
  - Major refactoring/improvements to RDKIT grid featurizer (879)
  - Installation/Testing/Docs improvements (892, 904, 907, 909, 920, 923)

1.3.0

This major version release consolidates around `TensorGraph` as DeepChem's high-level deep learning API of choice. Lots of improvements and bugfixes have been made to the core `TensorGraph` library, and many new layers and models have been added. In particular, DeepChem now features GANs, Seq2Seq models, Model Agnostic Meta Learning and more! Many improvements to tutorials, examples, website, and installation have been added as well.
  
  Our thanks to all the developers who contributed to this release, with a particular shout-out to those who made their first PRs to DeepChem!
  
  Detailed Changes:
  - TensorGraph Improvements (693, 705, 723, 730, 731, 746, 751, 753, 758, 760, 763, 766, 774, 778, 782, 783, 788, 791, 794, 799,  811, 817, 822, 824, 826, 833, 847, 850, 853, 854, 860, 871)
  - RL Improvements: Hindsight Experience Replay, Proximal Policy Optimization, API cleanup (686, 688, 697, 713, 719,  729,  740)
  - IPython Notebook Improvements (703, 706, 709, 711, 717, 721, 727, 745, 750, 829)
  - Cleaning up examples (755, 816, 819, 830, 840)
  - MoleculeNet Improvements (696, 718, 772, 854, 880, 738)
  - Rehaul Website (800, 801, 806, 807)
  - Improvements and extensions to `dc.splits` (690, 765, 784)
  - Improvements to rdkit-grid-featurizer (868, 873, 883)
  - Miscellaneous cleanups/fixes (701, 712, 724, 732, 735, 739, 796, 848, 885,
  - DeepChem installation/import improvements (737, 793, 802, 803, 804, 813, 814, 815, 852, 857, 859, 885, 888)
  
  Detailed listing of new models added:
  
  - GANs (855, 866)
  - Model Agnostic Meta Learning (759)
  - Seq2seq models (828)
  - ANI-1 Models (823, 839)
  - Spatial Filtering Graph Convs (851)
  - Message Passing Neural Networks (710)
  - TextCNNs (874)
  - Sluice Networks (805)

1.2.0

The major new change in this release is a new reinforcement learning framework. There have also been many upgrades and bugfixes to TensorGraph, large upgrades to MoleculeNet, and significant effort spent cleaning up and solidifying our test suite, documentation, and community standards.
  
  Detailed Changes:
  - MoleculeNet Upgrades (556, 578, 589, 592, 594, 606,  628, 629, 633, 661, 665, 667 )
  - Reinforcement Learning Support (557, 618, 640, 652, 656)
  - Cleaning up tests and making robust (560, 561, 563, 566, 582, 584, 595, 611, 662)
  - GPU Docker support (574)
  - Documentation Improvements (575, 597, 609, 615, 641, 670, 674)
  - TensorGraph Refinements and Debugging (552, 567, 569, 577,  603, 608, 636, 637, 638, 642, 655, 666)
  - Added code of conduct (580)
  - PyTorch Model Upgrades (646, 652)

1.1.0

This minor release version adds `TensorGraph`, a new backend for DeepChem models inspired by Keras's functional API. `TensorGraph` should now be ready for early users to start experimenting with. Over the next few releases, we will deprecate non-`TensorGraph` models in favor of the newer implementations. This release also contains a number of major improvements to MoleculeNet, with new models, datasets, and metrics.
  
  Detailed Changes:
  
  - Added a `contrib/` folder to allow users to contribute models/code more easily.
  - PyTorch multitask models merged into `contrib/` (481)
  - XGBoost models added (483)
  - A number of new datasets integrated with `dc.molnet` (484)
  - Basic queue support added . Allows for higher GPU utilization to be achieved (488)
  - Introduction of `TensorGraph`, a new backend for DeepChem inspired by Keras's function API (493, 505, 517, 520, 523, 544, 546 )
  - Adding support for weight-sharing between layers in `TensorGraph` (521)
  - Porting of multitask classifiers/regressors into `TensorGraph` and adding Dropout support (522)
  - Implementation of molecular DAG models (495)
  - Implementation of weave convolutions (496)
  - Adding atomic convolutions (509, 537 )
  - MPNNs added to `contrib/` (512)
  - Made `dc.splits` work with `NumpyDataset` (513)
  - Adding auPRC metric (531)

0.0.4

Made home-page point to git.

0.0.3

Moving to pbr to handle data files better in build.

0.0.2

Fixes a small issue in setup.py.

0.0.1

Preliminary release for build-automation purposes. Not yet a stable API.