Skip to main content
. 2023 Oct 23;3(11):100621. doi: 10.1016/j.crmeth.2023.100621

Table 2.

The comparison of Mol-GDL with SOTAs on six commonly used datasets, which contain only region tasks on molecular properties

RMSE MAE
Dataset ESOL FreeSolv Lipo QM7 QM8 QM9
No. molecules 1,128 642 4,200 6,830 21,786 133,885
No. average atoms 26(13) 18(7) 49(15) 16(3) 16(3) 18(3)
No. tasks 1 1 1 1 12 12
D-MPNN36 1.050(0.008) 2.082(0.082) 0.683(0.016) 103.5(8.6) 0.0190(0.0001) 0.00814(0.00001)
AttentiveFP37 0.877(0.029) 2.073(0.183) 0.721(0.001) 72.0(2.7) 0.0179(0.0001) 0.00812(0.00001)
NGramRF38 1.074(0.107) 2.688(0.085) 0.812(0.028) 92.8(4.0) 0.0236(0.0006) 0.01037(0.00016)
NGramXGB38 1.083(0.082) 5.061(0.744) 2.072(0.030) 81.9(1.9) 0.0215(0.0005) 0.00964(0.00031)
PretrainGNN39 1.100(0.006) 2.764(0.002) 0.739(0.003) 113.2(0.6) 0.0200(0.0001) 0.00922(0.00004)
GROVEbase40 0.983(0.090) 2.176(0.052) 0.817(0.008) 94.5(3.8) 0.0218(0.0004) 0.00984(0.00055)
GROVElarge40 0.895(0.017) 2.272(0.051) 0.823(0.010) 92.0(0.9) 0.0224(0.0003) 0.00986(0.00025)
GEM41 0.798(0.029) 1.877(0.094) 0.660(0.008) 58.9(0.8) 0.0171(0.0001) 0.00746(0.00001)
Mol-GDL 0.798(0.024) 1.809(0.100) 0.779(0.007) 62.2(0.4) 0.0205(0.0001) 0.00952(0.00013)

Note that the subindex indicates standard deviation values. Bolding indicates best results.