The idea behind all types of affinity labelling is that a molecule is constructed which resembles the substrate of an enzyme or other protein, and therefore binds to the target site noncovalently, by the normal processes of molecular recognition.