Skip to main content

Table 2 Stochastic of the applied dataset

From: DrugRep-HeSiaGraph: when heterogenous siamese neural network meets knowledge graphs for drug repurposing

Relationship

Entity

Node types

Relationship type

Database

The node numbers

The number of relationships

Intra-relationships

Drug

Name (\(\varphi\))

–

DrugBank [35]

410

–

Chemical Structure (\(C\))

has_chemical_substructure

PubChem [31]

881

52,979

Target Protein (\(T\))

has_target

has_interaction

DrugBank [35]

STRING [36]

1506

2122

6909

Protein Domain (\(D\))

has_domain

has_domain_target

UniProt [37]

1070

1828

2804

Side Effect (\(S\))

has_side_effect

SIDER4.1 [38]

5734

64,121

ATC Code (\(A\))

has_ATC_code

SIDER4.1 [38]

1087

2958

Disease

Name (\(\rho\))

–

DisGeNET [39]

141

–

Type (\({\mathbb{T}}\))

has_type

DisGeNET [39]

3

141

Class (\({\mathbb{C}}\))

has_class

DisGeNET [39]

22

9440

Gene (\({\mathbb{G}}\))

has_gene

has-gene-interaction

DisGeNET [39]

STRING [36]

3561

5891

16,870

Semantic (\({\mathbb{S}}\))

has-semantic-type

DisGeNET [39]

5

141

Inter-relationships

Target-Gene

has_encoded

UniProt [37]

 

633

Drug-Disease

has_treatment

repoDB [40]

 

748