Linking chemical and biological space for tracking target innovation trends
Barbara Zdrazil1, Lars Richter1, Nathan Brown2
1Unviversity of Vienna, Department of Pharmaceutical Chemistry
2BenevolentAI, London, UK
Preclinical data can guide decision making at earlier stages in the drug discovery pipeline since it is reflecting trends in research attention that the community is following over the years. It might help researchers to base their research endeavors on evidence-based criteria and direct their interest towards understudied proteins or core scaffolds (1). In this context, an emerging application of protein target indexing and characterization in drug discovery is the desire to capture the past, current, and (potentially) future degree of research interest in main drug target families (2). Up to now, explorative studies on research attention across target families have been carried out mainly from a static point of view with only a few analyses including the time dimension (3, 4).
In this study we were tracking compound-target associations over time in a target family-wise manner which delivered a picture of trends in research attention that the drug discovery community was following over time. With the aim to capture the chemical dimension of protein innovation trends, we explored differences in time trends of physicochemical compound properties across major target families. In addition, target innovation trends were linked to biological/therapeutic innovation patterns by linking targets to Gene Ontology (GO) and disease annotations within respective protein families. This biological dimension of target innovation was captured by fitting robust regression lines onto counts of measurements with respective GO terms or diseases, as well as in the form of dynamic network representations (capturing three different time spans) linking protein targets, GO annotations and diseases.
Interestingly, we could observe steep positive GO trends only for kinases, ion channels, and GPCRs, whereas proteases and nuclear receptors delivered only steep negative GO trends. Inspecting the connectivity of GO terms, protein targets and diseases in network representations over time revealed interesting trends, such as a decreasing interest in GPCR targets being responsible for “circulatory system processes” while at the same an increase in GPCR targets being involved in “immune system processes”.
(1) Zdrazil B, Guha R (2017) The Rise and Fall of a Scaffold: A Trend Analysis of Scaffolds in the Medicinal Chemistry Literature. J Med Chem.
(2) Oprea TI, Bologa CG, Brunak S, et al (2018) Unexplored therapeutic opportunities in the human genome. Nat Rev Drug Discov 17:317
(3) Rask-Andersen M, Almén MS, Schiöth HB (2011) Trends in the exploitation of novel drug targets. Nat Rev Drug Discov 10:579
(4) Santos R, Ursu O, Gaulton A, et al (2017) A comprehensive map of molecular drug targets. Nat Rev Drug Discov 16:19–34. https://doi.org/10.1038/nrd.2016.230
We acknowledge fruitful discussion with Rajarshi Guha (Vertex Pharm.).
