Skip to main content


Fig. 11 | Journal of Cheminformatics

Fig. 11

From: Machine learning models for hydrogen bond donor and acceptor strengths using large and diverse training data generated by first-principles interaction free energies

Fig. 11

Application example of the trained ML models using the best-performing radial atomic activity descriptors on four molecules. Acceptor and donor atoms that could participate in an internal hydrogen bond are marked in bold and the predicted hydrogen bonding strengths in kJ mol−1 for the respective atoms are displayed next to them. The experimental determinations of whether an internal hydrogen bond was formed or not were performed by NMR spectroscopy [86,87,88]

Back to article page