Fig. 4

Comparison of structural diversities between VQM24 and 4 other commonly used QM datasets, QM7b2, QM93, QM7-X12, ANI-1x11. (A) (Top) Scatter plots of normalized ratios of principal moments of inertia (NPMI) for all molecules from the 5 datasets. Title includes number of molecules within each dataset in brackets with k and M indicating thousand and million respectively. Rod, disc and sphere indicate NPMI values corresponding to linear, flat and spherical systems respectively. (B) (Middle) Histograms binning molecules by the number of heavy atoms (non-Hydrogen). (C) (Bottom) Bar plots indicating number of unique stoichiometries.