Deep potential model and dataset for water autoionization and glycine tautomerism in bulk water.

This dataset contained systems for a glycine molecule (with the zwitterionic, neutral, anionic, or cationic form) in water and pure water, labeled by the M06-2X functional. All systems include 0-1 pairs of hydronium and hydroxide ions, utilizing enhanced sampling for water autoionization.
This deep potential model can describe glycine tautomerism in water and water autoionization with M06-2X (hybrid meta-GGA functional) precision. The compressed model is more efficient for conducting molecular dynamics. For more details, see the paper and supporting information.