Knowledge Hub.

Working with NovaData.

The pharmaceutical industry is a major global market. According to IQVIA, in 2018, the global market for pharmaceuticals was $1.2 trillion. It is set to hit $1.5 trillion by 2023. There is an ever-growing need to develop drugs for existing and evolving diseases. Part of this process includes identifying drug candidates. A drug candidate is defined as ‘the molecule among several that has been shown to have sufficient target selectivity and potency, and favourable medicine-like properties and justifies further development’. This process can take 3-6 years.

Once identified, it will then be subjected to a new series of tests, non-clinical studies and clinical trials before it can become (hopefully) a medicine. The complete process can often take a very long period of time- ten years on average.

scientist-holding-test-tube-3912471-scaled

NovaData Solutions, a startup company that specialises in molecular modelling services, created an artificial intelligence platform to simulate the identification of drug candidates. This will significantly cut the time it takes to identify drug candidates and speed up the process for developing new medicines. NovaData’s approach is built on mature opensource libraries such as KNIME and TensorFlow.

NovaData used OCC to accelerate the development of the KNIME functionality required by their AI-based drug discovery process. NovaData could confidently use this to demonstrate the sort of solutions they are able to provide and generate publicity through open sourcing the node. “Oxford Computer Consultants picked up complex requirements quickly and were responsive and easy to work with,” said Mike Mazanetz, Director of Novadata Solutions.

OCC delivered a multi-parameter optimization node which is used in finding the best drug candidates. The OCC design team also lead the process of creating a polished and intuitive UX in collaboration with domain expertise from NovaData. The OCC design team managed this complete process end-to-end, from initial user research through to producing a final design and style guide.

OCC’s focus on quality meant we already had the processes and infrastructure required to satisfy the requirements for NovaData to release the MPO node as an open source community contribution.

Now that OCC has delivered the source code and related documentation material for the MPO node, NovaData may look to publish the node as a Trusted Community Extension. They may also make the code open source via GitHub or a similar platform.