A dataset containing 24 000+ entries of acid dissociation constant (p K a ) measurements was digitized and made publicly available, alongside a machine learning model for predicting macroscopic p K a values.
Zheng et al. (2026) studied this question.