. 2019 Jul 30;59(8):3370–3388. doi: 10.1021/acs.jcim.9b00237

Table 8. Number of Public Data Sets Where D-MPNN Is Statistically Significantly Better than, Equivalent to, or Worse than Each Baseline Model.

baseline	D-MPNN is better	D-MPNN is the same	D-MPNN is worse	no. of data sets
MoleculeNet²	5	3	2	10
Mayr et al.¹²	8	10	1	19
RF on Morgan	14	0	1	15
FFN on Morgan	14	5	0	19
FFN on Morgan Counts	15	4	0	19
FFN on RDKit	8	5	4	19