Digitized dataset of aqueous acid dissociation constants

Abstract

The acid dissociation constant (pK_a) quantifies the acidity of a compound, which is crucial for applications including drug design, environmental fate studies, and chemical synthesis. However, high-quality open-source digital pK_a datasets are scarce, which limits the ability for researchers to search for properties of individual compounds, while also limiting the potential of data-driven predictive models. In this work, we release the IUPAC Digitized pK_a Dataset, a digital version of a critically-assessed collection of data compiled up to 1970. The dataset includes metadata such as temperature, measurement method, assessed reliability of data, and chemical identifiers such as SMILES and InChI strings. The dataset spans 24 222 entries across 10 564 unique molecules, making it the largest FAIR open-source dataset publicly available for aqueous pK_a data. Herein, we detail the data digitization and checking process, and assess the informational space spanned by the data. We compare the new digital dataset to other widely-used datasets. Several pK_a predictors have been trained using these other datasets, but often have not been reliably tested due to overlap between the training and test data. We use the data to train a macroscopic pK_a predictor and determine its accuracy using overlap-free test data. The full dataset is available at https://doi.org/10.5281/zenodo.7236452, and the models and data splits used in this study are available at https://doi.org/10.5281/zenodo.18165948.