Skip to main content

Table 1 The first and second columns show each subset of drug features and the number of drugs which these features are available in the database, respectively. The third column indicates the number of drug-disease associations where the features are available in the database and the fourth one identifies the number of unknown drug-disease associations

From: The assessment of efficient representation of drug features using deep learning for drug repositioning

Subset ENo. of DrugsNo. of drug associations (\( \mathcal{A} \))No. of unknown drug-disease associationAvg. of AUCAUC
{s}424013,91625,235,2840.9420.944
{g}72961754,335,0200.8940.888
{e}67110,9503,984,8550.9270.926
{p}623316,84637,100,6690.9420.943
{g,s}72961754,335,0200.9360.933
{e,s}47183982,796,4070.8700.871
{s,p}322613,15919,197,6710.9410.941
{e,g}1554065918,9600.8560.844
{g,p}33759282,000,9070.8570.864
{e,p}60010,3053,562,6950.9090.906
{e,g,s}1554065918,9600.8490.848
{g,p,s}33759282,000,9070.8760.877
{g,p,e}1463944865,4860.8340.844
{e,p,s}44081622,612,0380.8680.870
{s,e,g,p}1463944865,4860.8400.846