Multiplex interaction prediction on DGIdb (transductive) benchmark leaderboard · SOTA2 Research

Multiplex interaction prediction on DGIdb (transductive)

71.5AUROCCAZI-MBN

Evaluation Results

Method	Links
CAZI-MBN2026.02		71.5	72.9	68.7	68.4
HDMI2026.02		55.1	55.7	54	51.1
DGCL2026.02		51.9	53.1	52.7	51.2
Graph Transformer2026.02		50.5	51.4	49.3	50.8
CAZI-MBN2026.02		0.715	0.729	0.687	0.684
HDMI2026.02		0.551	0.557	0.54	0.511
DMGI2026.02		0.547	0.542	0.533	0.505
R-GAT2026.02		0.544	0.549	0.521	0.514
R-GCN2026.02		0.536	0.541	0.53	0.508
MultiplexSAGE2026.02		0.532	0.539	0.549	0.507
DGCL2026.02		0.519	0.531	0.527	0.512
XGB (LLM)2026.02		0.517	0.572	0.541	0.488
CoSMIG2026.02		0.517	0.523	0.522	0.509
MLP (LLM)2026.02		0.511	0.561	0.536	0.503
Graph Transformer2026.02		0.505	0.514	0.493	0.508
DGI2026.02		0.503	0.507	0.51	0.5
GCN2026.02		0.482	0.493	0.511	0.491
GraphSAGE2026.02		0.447	0.467	0.527	0.473
XGB2026.02		0.35	0.389	0.579	0.421
MLP2026.02		0.284	0.375	0.609	0.407