Enzymes: Extracting Biological Insight Using UniProt

[@bamh1re318]

Great presentation, thanks! How many of the ~5,000 human enzymes have been targeted with clinical medicines? Is Lipinski's 'rule of five' still valid? He claimed there are ~500 enzymes considered "bottleneck" in biologically high impact nodules.

[@sib-swissinstituteofbioinf4579]

UniProt is probably not the first place to start looking for drug targets, and to obtain such numbers, as this is outside our field of expertise. But there are various resources that are trying to address drug targets and which are worth exploring, although they probably all define targets and druggability slightly differently: - IDG resources (Illuminating the Druggable Genome): They assign 4 different "target development levels" to drug targets: Tclin, Tchem, Tbio or Tdark. Tclin are targets for which an approved drug exists; Tchem are proteins that are not Tclin, but are known to bind small molecules with high potency, etc... (academic.oup.com/nar/article/51/D1/D1405/6851109). The PHAROS interface can be used to query the IDG data, e.g. for Tclin and Tchem pharos.nih.gov/targets?facet=Target%2BDevelopment%2BLevel!Tclin&facet=Target%2BDevelopment%2BLevel!Tchem , and to further filter this for enzymes. - DrugBank allows filtering for targets (e.g.go.drugbank.com/releases/latest#target-sequences) or enzymes. - OpenTargets www.opentargets.org/ - Human Protein Atlas (HPA) www.proteinatlas.org/humanproteome/tissue/druggable UniProtKB has cross-references to PHAROS, DrugBank, OpenTargets and HPA, and you can explore narrowing down the set of human enzymes to those entries with cross-references to one or more of these databases. Please don't hesitate to contact the UniProt helpdesk (www.uniprot.org/contact) with questions about UniProt, or directly write to the providers of the resources listed above. The UniProt team