The client’s project was investigating poly-pharmacology effects of a combination of antagonism at two or more Class A GPCRs. MedChemica examined the data from the literature and patents, performing both matched molecular pair analysis (MMPA), and model building to predict potency, based on random forest modelling of the unique fragments generated by the MMPA. Using the analysis key groups were identified where potency was gained at the GPCR targets, and critically where changes could be made. MCPairs Compounds-From-Rules generated ideas, and with further ideas from project medicinal chemists, were tested in-silico with the potency prediction models. Reduction to practice by the synthesis of these molecules generated novel compounds, and actives, to test in-vivo for the desired pharmacology. The diagram above illustrates the process and plots where key side chains were highlighted.
Indole-3-glyoxylamide-Based Series of Tubulin Polymerization Inhibitors
Researchers at the University of Sheffield wished to produce quality molecules possessing both in-vitro activity at the primary target and sufficient aqueous solubility and the correct metabolism profile. A small set of compounds were selected from the output of MCPairs Online Compounds-From-Rules AI engine where solubility was selected as the Goal. Synthesis of these led to compound 59 (see diagram) with sufficient aqueous solubility to formulate, and dose, in a mouse xenograft study and show efficiency. The results were published in 2015 and further investigations are on-going.
Custom interfacing to MCPairs Enterprise
A corporate user of MCPairs Enterprise loaded and processed their own compounds and measurements, and produced their own unique database of knowledge. Although the MCPairs Visualiser (GUI or Webtool see diagram) was available, behind their firewalls, for chemists to use, they wished to directly access the tools within their own in-house design tools. MCPairs Enterprise is a REST API driven server and can be linked by standard POST, PUT, GET and DELETE requests with data supplied in JSON format. The corporate user engaged a third party software engineer to join their own software tools via the REST API – the work took 3 weeks to complete. In a similar fashion Optibrium have attached StarDrop to MCPairsOnline and MedChemica can provide simple Knime nodes to develop workflows.