Table 3. Open in new tab Performance... | Oxford Academic

Table 3.

Open in new tab

Performance comparison on classification (measured by ROC-AUC and regression [measured by RMSE] tasks).^a

Method	ESOL $↓$	FreeSolv $↓$	Lipop $↓$	Average	BACE $↑$	BBBP $↑$	ClinTox $↑$	Tox21 $↑$	SIDER $↑$	Average
GCN	0.778	1.582	0.899	1.086	0.829	0.895	0.615	0.788	0.624	0.750
GIN	0.619	1.136	0.756	0.837	0.850	0.890	0.753	0.824	0.618	0.787
GROVER	0.983	2.176	0.817	1.325	0.826	0.700	0.812	0.743	0.648	0.746
GraphMvp	1.092	^b	0.681	^b	0.812	0.724	0.790	0.631	0.639	0.719
GEM	0.798	1.877	0.660	1.109	0.856	0.724	0.906	0.781	0.672	0.788
GeomGcl	0.764	0.877	0.544	0.728	^b	^b	0.917	0.851	0.647	^b
HiMol	0.833	2.283	0.708	1.275	0.846	0.732	0.808	0.762	0.625	0.755
3D PGT	1.061	^b	0.687	^b	0.809	0.721	0.794	0.738	0.606	0.734
Uni-miol	0.788	1.48	0.603	0.957	0.857	0.729	0.919	0.796	0.659	0.792
3DGCL	0.778	1.441	^b	^b	0.792	0.855	^b	^b	^b	^b
MolPath(ours)	0.266	0.200	0.524	0.330	0.898	0.913	0.863	0.868	0.674	0.843

Method	ESOL $↓$	FreeSolv $↓$	Lipop $↓$	Average	BACE $↑$	BBBP $↑$	ClinTox $↑$	Tox21 $↑$	SIDER $↑$	Average
GCN	0.778	1.582	0.899	1.086	0.829	0.895	0.615	0.788	0.624	0.750
GIN	0.619	1.136	0.756	0.837	0.850	0.890	0.753	0.824	0.618	0.787
GROVER	0.983	2.176	0.817	1.325	0.826	0.700	0.812	0.743	0.648	0.746
GraphMvp	1.092	^b	0.681	^b	0.812	0.724	0.790	0.631	0.639	0.719
GEM	0.798	1.877	0.660	1.109	0.856	0.724	0.906	0.781	0.672	0.788
GeomGcl	0.764	0.877	0.544	0.728	^b	^b	0.917	0.851	0.647	^b
HiMol	0.833	2.283	0.708	1.275	0.846	0.732	0.808	0.762	0.625	0.755
3D PGT	1.061	^b	0.687	^b	0.809	0.721	0.794	0.738	0.606	0.734
Uni-miol	0.788	1.48	0.603	0.957	0.857	0.729	0.919	0.796	0.659	0.792
3DGCL	0.778	1.441	^b	^b	0.792	0.855	^b	^b	^b	^b
MolPath(ours)	0.266	0.200	0.524	0.330	0.898	0.913	0.863	0.868	0.674	0.843

The SOTA results are shown in bold. The underlined is the second best.

The result is unavailable in the original paper.

Table 3.

Open in new tab

Performance comparison on classification (measured by ROC-AUC and regression [measured by RMSE] tasks).^a

Method	ESOL $↓$	FreeSolv $↓$	Lipop $↓$	Average	BACE $↑$	BBBP $↑$	ClinTox $↑$	Tox21 $↑$	SIDER $↑$	Average
GCN	0.778	1.582	0.899	1.086	0.829	0.895	0.615	0.788	0.624	0.750
GIN	0.619	1.136	0.756	0.837	0.850	0.890	0.753	0.824	0.618	0.787
GROVER	0.983	2.176	0.817	1.325	0.826	0.700	0.812	0.743	0.648	0.746
GraphMvp	1.092	^b	0.681	^b	0.812	0.724	0.790	0.631	0.639	0.719
GEM	0.798	1.877	0.660	1.109	0.856	0.724	0.906	0.781	0.672	0.788
GeomGcl	0.764	0.877	0.544	0.728	^b	^b	0.917	0.851	0.647	^b
HiMol	0.833	2.283	0.708	1.275	0.846	0.732	0.808	0.762	0.625	0.755
3D PGT	1.061	^b	0.687	^b	0.809	0.721	0.794	0.738	0.606	0.734
Uni-miol	0.788	1.48	0.603	0.957	0.857	0.729	0.919	0.796	0.659	0.792
3DGCL	0.778	1.441	^b	^b	0.792	0.855	^b	^b	^b	^b
MolPath(ours)	0.266	0.200	0.524	0.330	0.898	0.913	0.863	0.868	0.674	0.843

Method	ESOL $↓$	FreeSolv $↓$	Lipop $↓$	Average	BACE $↑$	BBBP $↑$	ClinTox $↑$	Tox21 $↑$	SIDER $↑$	Average
GCN	0.778	1.582	0.899	1.086	0.829	0.895	0.615	0.788	0.624	0.750
GIN	0.619	1.136	0.756	0.837	0.850	0.890	0.753	0.824	0.618	0.787
GROVER	0.983	2.176	0.817	1.325	0.826	0.700	0.812	0.743	0.648	0.746
GraphMvp	1.092	^b	0.681	^b	0.812	0.724	0.790	0.631	0.639	0.719
GEM	0.798	1.877	0.660	1.109	0.856	0.724	0.906	0.781	0.672	0.788
GeomGcl	0.764	0.877	0.544	0.728	^b	^b	0.917	0.851	0.647	^b
HiMol	0.833	2.283	0.708	1.275	0.846	0.732	0.808	0.762	0.625	0.755
3D PGT	1.061	^b	0.687	^b	0.809	0.721	0.794	0.738	0.606	0.734
Uni-miol	0.788	1.48	0.603	0.957	0.857	0.729	0.919	0.796	0.659	0.792
3DGCL	0.778	1.441	^b	^b	0.792	0.855	^b	^b	^b	^b
MolPath(ours)	0.266	0.200	0.524	0.330	0.898	0.913	0.863	0.868	0.674	0.843

The SOTA results are shown in bold. The underlined is the second best.

The result is unavailable in the original paper.