ILThermoPy is a Python library to access ILThermo 2.0, which is the NIST standard reference database, containing measured physico-chemical properties for a wide spectrum of ionic liquids. Despite the significant amount of stored data, ILThermo 2.0 cannot be used for machine learning, parametrization of empirical physical models, and other data-driven approaches as is. The reason is that ILThermo 2.0 describes molecular structure of IL's components with a chemical name only, whereas SMILES, InChI and other structure identifiers are not available.
ILThermoPy solves this problem via the preliminary semi-automatic conversion of compound names to SMILES with subsequent manual validation. This allows one to conduct a substructural search and to immediately generate chemoinformatic descriptors for the extracted data.
-
There is no official web API available to access ILThermo 2.0, therefore the stability of this library depends on the stability of the JSON/javascript framework of the webservice.
-
ILThermo 2.0 database is regularly updated, at least once a year. Those updates change internal compound IDs, which are used to add SMILES to the extracted data. Therefore, after database's update old versions of ILThermoPy can fail to retrieve structural data at least for some of the new entries.
ILThermoPy can be installed via PyPI:
> pip install ilthermopy
-
Python 3.7+;
-
requests;
-
pandas;
-
importlib_resources (for Python 3.7 and 3.8).
-
ILThermo 2.0: webapp accessing ILThermo 2.0 database.
-
PyPI package: PyPI page of the package.
-
Documentation: cookbook, descriptive API, and other useful information.
-
Update tools: script for semi-automatic update of structural information of new ILThermo compounds after database's update.