Skip to main content


Table 1 Per-atom input features computed from RDKit molecules

From: Rapid prediction of NMR spectral properties with quantified uncertainty

Feature Description Number
Atomic number   1
Atomic number One-hot encoded {H, C, O, N, P, S, F, Cl} 8
Valence   1
Valence One-hot encoded 1–6 6
Aromaticity Does RDKit identify this atom as being part of an aromatic structure 1
Hybridization state One-hot encoded s, sp, \(sp^2\), \(sp^3\), \(sp^3d\), \(sp^3d^2\) 6
Formal charge Does this atom have a net charge, one-hot encoded {\(-\,1\), 0, \(+\,1\)} 3
Default valence One-hot encoded 1–6 6
Rings Is this atom the member of a ring , one-hot encoded 3–7 5
Total   37