The American biotech Polarisqb have partnered with the tech giant Fujitsu to combine pioneering quantum-inspired technology, machine learning, hybrid quantum mechanics and molecular mechanics simulations (QM/MM), in a novel small molecule discovery platform.
This technology will significantly improve the speed, efficiency, and chemical diversity of molecule leads returned, the companies announced in a press release last week. Initially the technology will be applied to designing lead candidates targeting Dengue virus, which currently has no approved therapies.
Fujitsu are contributing their quantum-inspired Digital Annealer which enables the platform to scan an exponentially larger molecule repertoire and solve combinatorial optimisation problems up to 10,000x faster than current market solutions*. Lead molecules identified by the Digital Annealer will then be refined by the Polarisqb combined proprietary ML algorithms and QM/MM simulations to rapidly assess drug viability. Specifically, the ML algorithms rank physiochemical properties i.e. ADMET predictions of the leads generated, whilst QM/MM simulations evaluate binding affinities. The resulting short-list of molecules are prioritised for synthesis and in vitro testing.
This platform could cut the time typically taken for molecule identification through to the pre-clinical development from 2-4 years to just eight months. Both parties hope that a short-list of novel molecules will be available to share with stakeholders by May 2020, targeting a previously un-drugged Dengue virus protein.
CEO of Polarisqb, Dr Shahar Keinan, said “Polarisqb is excited to be co-creating this novel platform that can revolutionise drug design. We believe drug development has reached an inflection point, and that leveraging advances in quantum-inspired computing will dramatically increase the speed of computation and diversity of candidate molecules, answering clear needs in the market.”
