Skip to main content

Table 5 Descriptions used for the ChEMBL confidence scores

From: Target prediction utilising negative bioactivity data covering large chemical space

Confidence score Description
0 Default value—target assignment has yet to be curated
1 Target assigned is non-molecular
3 Target assigned is molecular non-protein target
4 Multiple homologous protein targets may be assigned
5 Multiple direct protein targets may be assigned
6 Homologous protein complex subunits assigned
7 Direct protein complex subunits assigned
8 Homologous single protein target assigned
9 Direct single protein target assigned
  1. A confidence score of 5 was used in this study to ensure that multi-domain proteins are not excluded from training data. Confidence scores of 9 or 8 are frequently used which specify homologous or direct proteins to be assigneds