Skip to main content
. 2023 Jun 30;39(Suppl 1):i326–i336. doi: 10.1093/bioinformatics/btad222

Table 1.

Comparison with state-of-the-art approaches.

Method VS QED SA HAR (%) Diversity
Reference −7.550 (3.2e−2) 0.476 (7.7e−23) 3.453 (2.0e−5)
LiGANN −6.144 (2.5e−269) 0.371 (0.0) 4.787 (1.9e−88) 23.8 0.655
3D-SBDD −6.344 (1.2e−174) 0.502 (0.0) 3.912 (1.8e−39) 29.1 0.742
Pocket2Mol −7.288 (2.9e−92) 0.563 (6.6e−233) 3.205 (1.4e−7) 54.2 0.688
Transformer −7.385 (1.3e−33) 0.512 (1.6e−119) 2.756 (2.9e−11) 49.3 0.725
AlphaDrug −7.393 (1.2e−29) 0.507 (1.1e−118) 2.620 (3.1e−14) 50.1 0.727
Decoys −6.737 (2.1e−15) 0.539 (3.5e−66) 3.830 (5.0e−69) 29.6 0.739
CProMG −7.644 0.741 2.884 55.5 0.757

Note: Boldface values represent the best values of the metric.