Accurately predicting solubility curves via a thermodynamic cycle, machine learning, and solvent ensembles

The datasets and supplementary materials for the manuscript "Accurately predicting solubility curves via a thermodynamic cycle, machine learning, and solvent ensembles" by Emad Al Ibrahim, Nathan Morgan, Simon Muller, Saikiran Motati, and William H. Green. The files include compiled data for solubility, logP, enthalpy of fusion, melting point, and activity coefficients. *Note that only the data approved for public release is shared here. Description:dHfus_DB.csv: Enthalpy of fusion data - used for training/validationTmp_DB.csv: Melting point temperature data - used for training/validationDrug_like_molecules: Folder containing the drug-like solutes added to gamma_QM_DBgamma_QM_DB: Activity coefficient data from COSMO-RS - used for training/validationlogP: Folder containing logP datasets used for testing (OPERA, SAMPL6, and SAMPL7)Solubility: Folder containing solubility datasets used for testing - all_solubility.csv is the data from 15 datasets, and BiggerSolDB is the combination shared both with and without Zwitterions