Please refer to the git repository (add a link to https://github.com/fhvermei/SolProp_ML) for detailed installation guide, sample code, citation, and license information
General information: <\b>
Machine learning and thermodynamics for the prediction of solubility of pharmaceuticals in organic solvents and water. Package includes code and trained ML models.
Github code: https://github.com/fhvermei/SolProp_ML
Read & cite publication:
Web-interface: https://rmg.mit.edu/database/solvation/search/
To install: <\b>
conda install -c fhvermei -c conda-forge solprop_ml
For temperature dependent calculations, additional install required:
pip install git+https://github.com/bp-kelley/descriptastorus
To run (more info on github):<\b>
import solvation_predictor.calculate_solubility as calc
results = calc.calculate_solubility(path=path, df=df, validate_data_list=['solute', 'solvent', 'reference_solvent', 'temperature'], calculate_aqueous=True, calculate_Hdiss_T_dep=True, reduced_number=False, export_csv='./results_test.csv', export_detailed_csv=True, solv_crit_prop_dict=None, logger='./test.log')
