Classification of biological data

Biological data are represented by pChEMBL values with each entry labelled as either active or inactive.
DELTA contains 84400 biological data, covering 100 active and 100 inactive ligands for each target.
DELTA-X contains 288154 biological data, where the number of active and inactive molecules is defined by the minority class for each target.

1. Calculation of pChEMBL

pChEMBL is provided to ease the comparison of activity of different ligands, values are calculated as:
pChEMBL=-Log(molar IC50, XC50, EC50, AC50, Ki, Kd or Potency).
For additional information please see ChEMBL documentation.

2. Distinction between active and inactive ligands

To differentiate between active and inactive ligands, a threshold has been established at a pChEMBL value of 5.5:
ligands with a pChEMBL value higher than 5.5 are labelled as active (1), whereas those with a pChEMBL value lower than or equal to 5.5 are labelled as inactive (0)*.

*The reported pChEMBL values are rounded to two decimal places. If a pChEMBL value fell between 5.501 and 5.549, it is displayed as 5.5; however, the molecule remains classified as active.

Explore in Real-Time

Biological data for the full dataset can be accessed (for the DELTA dataset only) in the Explore Ligands page. For each molecule, the pChEMBL value and its active (1) or inactive (0) label are provided.

Biological data for a specific target can be accessed (for the DELTA dataset only) in the ligands page accessible from the ‘Ligands’ column of the considered target in the Explore Targets page. For each molecule, the pChEMBL value and its active (1) or inactive (0) label are provided.

Get the Full Data

DELTA

Biological data for the DELTA dataset can be downloaded for the entire database by clicking here. The CSV file contains the following information:

• Target ID: DELTA unique identifier for the target.
• Target Name: Official name of the target protein.
• Molecule Name: CHEMBL ID or PubChem CID.
• Synonyms: Other known identifiers.
• SMILES: SMILES string of the molecule.
• pChEMBL: pChEMBL value.
• Class: active (1) or inactive (0).

DELTA-X

Biological data for the DELTA-X dataset can be downloaded for the entire database by clicking here. The CSV file contains the following information:

• Target ID: DELTA unique identifier for the target.
• Target Name: Official name of the target protein.
• Molecule Name: CHEMBL ID or PubChem CID.
• SMILES: SMILES string of the molecule.
• pChEMBL: pChEMBL value.
• Class: active (1) or inactive (0).