On 19th September, 2019 FUJIFILM Diosynth Biotechnologies (FDB), in partnership with the Knowledge Transfer Network (KTN), held a “Science Exchange” event, inviting innovative companies to help transform the next generation of biologic drugs manufacture, through the application of advanced digital technologies. Intellegens was invited to introduce their technology to the bioprocessing and biotechnology industries.
This was an opportunity for businesses, spin-outs and academics to meet key FDB personnel, find out more about the company’s Innovation Strategy, and to form collaborations. The event was introduced by Andy Topping (CSO, FDB) and Andy Jones (Medicines Manufacturing Industrial Challenge Director). Pitching and information sessions were held throughout the day from companies and academics who are leading the way in new approaches to digitising the biopharma industry, followed by a panel discussion. The following report highlights the presentations, discussions and emerging themes from the day.
One of the challenges of using AI is that techniques require very clean and complete data. But experimental and simulation data can be sparse and noisy. The University of Cambridge spin-out, Intellegens, is using deep learning algorithms to train models on data as sparse as 0.05%, to help companies move away from a time consuming trial and improvement approach. Its system brings all the available data together, and uses underlying correlations to predict missing values.
The company initially worked on optimising Nickel superalloys for Rolls Royce – designing four new alloy families, which have been patented. Across various clients, its AI optimised approach has reduced prototype costs by around 80%, says Commercial Director Jamie Smith. Less materials wastage means a reduced environmental impact. “The deep learning approach we’ve developed is a data science approach rather than a physics based approach – so it enjoys as much success in drugs as in alloys.” It’s possible, says Smith, to ‘tweak’ the model – if experts know, for example, that property ‘x’ can’t be influenced by property ‘y’. Clients might not necessarily reduce the number of experiments they do, but use the results to guide where the next experiment should be.
A demonstration of its AI engine Alchemite can be found at https://app.intellegens.ai
Intellegens is a spin-out from the University of Cambridge with a unique Artificial Intelligence (AI) toolset that can train deep neural networks from sparse or noisy data. The technique, created at the Cavendish Laboratory, is encapsulated in Intellegens first commercial product, AlchemiteTM. The innovative deep learning algorithms that AlchemiteTM is based on can see correlations between all available parameters, both inputs and outputs, in fragmented, unstructured, corrupt or even noisy datasets. The result is accurate models that can predict missing values, find errors and optimise target properties. Capable of working with data that is as little as 0.05% complete, AlchemiteTM can unravel data problems that are not accessible to traditional deep learning approaches. Suitable for deployment across any kind of numeric dataset, AlchemiteTM is delivering ground breaking solutions in drug discovery, advanced materials, patient analytics and predictive maintenance – enabling organisations to break through data analysis bottlenecks, reduce the amount of time and money spent on research, and support better, faster decision-making. For more information contact us here.