AIRA: AI-based Research Assistant for Active Site Structure and Dynamics from Energy Landscapes

This project integrates computational and experimental approaches in a new cloud platform that provides access to state-of-the-art energy landscapes tools and AI-based assistance for biomolecular design and drug discovery. It brings together Dodrotu Limited, the Wales group at the University of Cambridge, and the Schnell group at the University of Oxford, to deliver an energy landscapes AI-based solution for discovery and analysis of influenza drug-binding sites and epitopes and explanation and characterisation of binding mechanisms, formulated in a way that allows for experimental validation with NMR and rapid prototyping of designs. Dodrotu is a UK information technology services startup established in 2019 and incubated at Oxford University Innovation. Dodrotu has developed a secure, end-to-end encrypted platform for delivery of applications that augment traditional communications with AI. The founders of Dodrotu were members of the Wales Group and have expertise in optimisation, machine learning, signal and natural language processing, artificial neural networks, functional programming, VoIP and mobile technologies. Professor David J. Wales is Chair of the Theory Group in the Chemistry Department at Cambridge University. He has developed new theory and computational methods for exploring energy landscapes in molecular and condensed matter science, which have been recognised by named lectures, visiting professorships, and awards in the UK and overseas, including election to the Royal Society in 2016\. Professor Jason R. Schnell is a group-leader in Oxford University's Department of Biochemistry and an expert in protein structure and dynamics. His group recently advanced our understanding of influenza, and has detailed knowledge of high entropy regions and mechanisms of polymerisation and membrane binding of Influenza A virus M1 protein. Our ambition is to provide new tools that will both greatly speed up drug discovery and lower its costs. After validation on influenza as part of this project, the new methodology will find a wide range of applications in molecular science, from small molecule targeting of key proteins in health and disease, to protein design for protein therapeutics and industrial enzymes.