The remit of our company, a charity and small biotech, is to help accelerate the process of target validation and generating tool compounds for enabling further exploration of new targets (as proposed by our various academic collaborators in the UK and across the world). Essentially this means we primarily focus on hit/lead discovery for challenging targets and hence we normally encounter programs where there is sparse structural information. In this webinar I will endeavour to describe how we use a combination of computational tools to drive these activities, focusing on our efforts to embed in silico tools (including FTrees and others) into our normal hit to lead discovery protocols.
And to exemplify I will present a recent project where, using a combination of available ligand-based methods, some homology modelling, solvent simulation (to model favourable water pockets), fragment docking and in silico directed mutagenesis we managed to help our chemists to deliver the first documented series of low nanomolar ATP competitive inhibitors of the human PAICS enzyme from low affinity fragments, a potential drug target for metastatic breast cancer. This was achieved with no x-ray structures. The series of novel inhibitors has displayed exquisite selectivity against human kinases, high activities against relevant cancer cells and yielded candidates with suitable ADMET and PK profiles, the best of which have been submitted for preliminary in vivo investigation against mouse Xenograft models.