A survey of the multi-target testing experimental practices inferred from the ChEMBL dataAshenafi Legehar1, Leo Ghemtio1, Henri Xhaard1 |
|
| 1Faculty of Pharmacy, Division of Pharmaceutical Chemistry and Technology, University of Helsinki | |
| Polypharmacology is regarded as extremely important in drug discovery (Reddy & Zhang, 2013), however little investigation has been conducted on the practices leading to its discovery. We thus conducted an extensive analysis on the data deposited in ChEMBL(Bender, 2010; Gaulton et al., 2017). We focus on four journals (J.Med.Chem, E.J.Med.Chem., Bioorg.Med.Chem., Bioorg.Med.Chem.Lett) that forms the bulk (90%) of the deposited compound data over the last 30 years. A set of simple keywords allow to focus on ~8000 articles dealing with synthetic medicinal chemistry and containing ~155000 compounds, most probably congeneric series, i.e. to exclude articles that are perspectives, review, QSAR models, large screens, etc. The number of compounds with registered activities is in the 20-26 range per article over journals and time periods; while the number of targets is in the 6-9 range. Note that this does not account for studies with no apparent targets, For example, antibacterial. The three main tested targets are not surprisingly kinases (13% of the compounds are tested on kinases, whereas kinases are listed as a target in 15% of the articles), GPCRs (30% and 28%), and serine proteases (15% and 13%). We furthermore show that metabolism (CYP450) and toxicity proteins (hERG) are often tested within the same article with these targets. Zooming into GPCRs show intra-family testing as well as monoamine transporters cross-targets, but much less testing of CYPs and hERG, presumably reflecting a longer distance to the clinic. Thus, compounds are not only screened multiple times as part of collections or for repurposing purposed; ADME testing, for example off-target screens (typically kinase panels and GPCR panels), CYP450 and hERG, also provide a considerable yet uncharacterized amount of biological activity data. This work is to be submitted to the special issue “Multi-Target Drug Discovery: an opportunity for novel and repurposed bioactive compounds” for the Mu.Ta.Lig COST action CA15135 http://www.mutalig.eu Bender, A. (2010). Compound bioactivities go public. Nature Chemical Biology, 6(5), 309–309. https://doi.org/10.1038/nchembio.354 |
