Change Log

v3.1.0 (2024-11-21)

  • Removed langchain and switched to use openai API directly

v3.0.3 (2023-06-19)

  • Now compatible with python 3.11

v3.0.2 (2023-02-23)

  • Now try to find largest component with multiple compounds separated via . in SMILES

  • Added langchain for text explanations

v3.0.1 (2023-02-02)

  • Removed ~ any bonds from SMILES strings coming from synspace

v3.0.0 (2023-01-24)

  • Added synspace method to generate synthetically feasible space

  • Added drug-like filter by default for counterfactuals

v2.2.1 (2022-12-7)

  • Fixed bug in sorting for text explanations

  • Fixed empty plot names saying None

  • Added priority for naming and removed invalid names

  • Added more names (metyhl, ethyl, butyl, etc)

  • Fixed sample_space to accept partials or objects

  • Added openai prompting

  • Added name_morgan_bit as external facing

v2.2.0 (2022-11-3)

  • Added natural language explanation method

v2.1.1 (2022-6-03)

  • Fixed plotting bug in plot_descriptors

v2.1.0 (2022-6-03)

  • plot_descriptors will no longer save to file without filename

v2.0.1 (2022-5-31)

  • Made default run_stoned argument use basic instead of semantically robust alphabet, as

    claimed in documentation

v2.0.0 (2022-5-31)

  • Added surrogate model explanation method

  • Added support for attributing ECFP, MACCS fingerprints, rdkit descriptors and plotting them

  • Example notebooks for new method

  • Fixed chirality stripping in sanitize

  • Made it possible to use multiple base molecules for ECFP descriptors

v1.1.0 (2022-5-2)

  • Made it possible to have model function that only takes one argument instead of SMILES and SELFIES

v1.0.2 (2022-5-2)

  • Switched to bulk Tanimoto to improve speed

  • Tightened chemed api limit

v1.0.1 (2022-4-4)

  • Added quiet mode

v1.0.0 (2022-1-21)

  • Switched to SELFIES v2.0 release, which has a different token list

  • Removed “experimental” tag from Chemed and Custom methods

  • Type annotations now pass mypy

  • Paper models now generate SVGs correctly and fixed token issues

v0.6.0 (2022-1-17)

  • Changed behavior of num_samples so that it is not affected by mutation count in STONED

v0.5.2 (2022-1-4)

  • Fixed SMILES escaping in URL in chemed

v0.5.1 (2021-11-16)

  • Fixed similarity float vs int in chemed

v0.5.0 (2021-10-26)

  • Added custom lists for counterfactual source (contributed by @maclandrol)

v0.4.1 (2021-09-27)

  • Fixed randint usage in stoned.py

v0.4.0 (2021-09-17)

  • Refactored code into files

  • Added SVG rewrite so mol structures are SVGs

  • SVGs are handled with skunks

  • Added cartoon style for scatter plot in plot_space

v0.3.2 (2021-09-02)

  • Added SMILES sanitization before generating sample space

v0.3.1 (2021-08-26)

  • Fixed num_samples not passed correctly

  • Fixed unparsable simles coming from pubchem

v0.3.0 (2021-08-25)

Added CHEMED method and progress bar

v0.2.0 (2021-08-14)

Initial package release