This dataset is from a selection of attributed target lists extracted from the literature as supplementary data for other downloadable databases. The criteria for inclusion are drug target coverage for these human proteins. However, the exact definition varies between lists, as explained in the description below. This includes different terminology (e.g. “successful, “approved” or “proven”). There are also differences in primary target (~1:1 drug: protein) vs. secondary or subunit mappings (1:many).
The dataset includes utilities that can be explored; there are two that might be considered: a) following the database links and b) comparing them for intersects (protein IDs in common) and differentials (protein IDs unique to particular lists or subsets). The information extends to comparisons with lists that can be generated in the course of other studies or other published work (e.g. expression data or disease association gene candidates). Other areas that can be explored are: a) what other utilities that are found valuable and b) other recently published target lists recommended for inclusion.
The metadata descriptions from this source are minimal since context is provided either in the references and/or the download descriptions for the appropriate databases or sources. The lists are Excel sheets of UniProtKB, HGNC and ChEMBL live links.
Lists that are not UniProtKB Accessions in the first place are normalized to these (e.g. mappings of Human Gene Nomenclature Committee (HGNC) Symbols or Entrez Gene IDs (EGID) to UniProtKB). They are then filtered to human and Swiss-Prot (i.e. any TrEMBL entries are removed) and to approved drug targets if this is an option in the original list. In such cases lists that are hosted thus become transformations, rather than direct facsimiles, of the primary sources.
Given such ID cross-mappings are not perfect; absolute correctness is not guaranteed. However, International Union of Basic and Clinical Pharmacology (IUPHAR) and the British Pharmacological Society (BPS) Guide to PHARMACOLOGY versions are supplied in good faith and the originals are available in every case.
If readers are unfamiliar with protein list “slicing and dicing” the source recommends the following:
1. The UniProtKB interface
2. Venny (for comparing up to four lists)
3. Panther (for displaying detailed protein classifications and attributes from lists)
“The largest innovations within pharmaceutical development come through new compounds that have unique and novel modes of action. These innovations commonly involve expanding the protein space targeted by pharmaceutical agents. At present, information about drugs and drug targets is available online via public databases such as DrugBank and the Therapeutic Targets Database. However, this information is biased, understandably so, toward established drugs and drug-target interactions. To gain a better overview of the drug-targeted portion of the human proteome and the directions of current drug development, we developed a data set of clinical trial drug-target interactions based on CenterWatch’s Drugs in Clinical Trials Database, one of the largest databases of its kind. Our curation identified 475 potentially novel clinical trial drug targets. This review aims to identify trends in drug development based on the potentially novel targets currently being explored in clinical trials.” PubMed Abstract on The druggable genome: Evaluation of drug targets in clinical trials suggests major shifts in molecular class and indication, Epub 2013 Aug 30.