In bindingDB documentation the value Ligand HET ID in PDB and PDB ID(s) for Ligand-Target Complex are defined as follows:
Ligand HET ID in PDB. If available, the hetero group ID for this ligand in the PDB.
PDB ID(s) for Ligand-Target Complex. If available. Criterion for protein match is 85% sequence
identity of protein and exact match for ligand.
This maps the ligand HET ID to PDBs that it appears (sometimes in several different protein complexes, including binding binding site mutants). Thus, the assigned value does not mean that the reported assay value refers to this exact PDB complex.
Unfortunately, plinder pipeline code use these PDB IDs provided in PDB ID(s) for Ligand-Target Complex. to match the binding affinity values, without further checks on the sequence:
https://github.com/plinder-org/plinder/blob/main/src/plinder/data/pipeline/transform.py#L161
As a result, the system_has_binding_affinity and ligand_binding_affinity values are incorrect and should not be used!
In bindingDB documentation the value
Ligand HET ID in PDBandPDB ID(s) for Ligand-Target Complexare defined as follows:This maps the ligand HET ID to PDBs that it appears (sometimes in several different protein complexes, including binding binding site mutants). Thus, the assigned value does not mean that the reported assay value refers to this exact PDB complex.
Unfortunately,
plinderpipeline code use these PDB IDs provided inPDB ID(s) for Ligand-Target Complex.to match the binding affinity values, without further checks on the sequence:https://github.com/plinder-org/plinder/blob/main/src/plinder/data/pipeline/transform.py#L161
As a result, the
system_has_binding_affinityandligand_binding_affinityvalues are incorrect and should not be used!