56491Combined AI and modelling service for rapid response drug designClosedFeasibility StudiesInnovate UKWhen new diseases such as COVID-19 emerge, it is not known whether existing drugs could be effective or new drugs need to be designed. The lack of information on an emerging disease makes these problems hard to solve, and it is made worse by the fact that subtle differences between diseases can have a large impact on which medicines are effective. In this project, we are using AI and computer simulations of drug and virus interactions for a known disease to find treatments for an emerging one, which shares similarities to the known disease. Additionally, when there is not enough data to decide this, our models will suggest the best potential drugs to be made and tested in order to quickly understand the different requirements of treating the new disease. Making the best choices is important as testing is both time consuming and expensive. In this project, we will focus on adapting models based on the SARS outbreak in 2002-2004 to search for drugs which might be effective against COVID-19. This approach is made possible because not only is there significant data from the related disease, but UK scientists rapidly conducted experiments showing how bits of drugs (known as 'fragments') bound to their targets within the new COVID virus. This early information provides an excellent starting point for comparing the two viruses, enabling the use of existing data to find new treatments. This work will build on Kuano's existing AI and simulation platform which was developed to design new cancer treatments. The technology developed will be applicable to future disease outbreaks as well as being able to inform more long term drug discovery projects. Successful development of our platform means that in addition to our initial goals we will make and test a small number of potential anti-COVID drugs designed by our platform. In order to bring our platform to market, we will also create a dashboard to communicate how our model integrated different data to efficiently create more diverse chemistry to potential clients.