Intellegens, AI company with unique deep learning toolset for sparse data, and Optibrium™, leading providers of software and services for drug discovery, today announced they have reached a further significant milestone in their contribution to the Open Source Malaria (OSM) initiative.
Intellegens and Optibrium achieve further success in Open Source Malaria initiative
The team has successfully completed phase 2 of a global challenge aimed at developing and testing novel antimalarial compounds. During this phase, predictive models from phase 1 were combined with generative methods to design novel compounds. The compounds were subsequently validated by testing their activity against the target. Out of four compounds proposed in this phase, only Optibrium/Intellegens’ entry demonstrated potency against the target, indicating the powerful combination of StarDrop™, Optibrium’s computational platform for small molecule design and optimisation, with the AI-powered (Alchemite™) technologies of its Augmented Chemistry™ platform.
In the latest phase of the OSM project, the team deployed the in silico, generative chemistry capabilities of StarDrop™ to design new compounds predicted to be active against a putative target in Plasmodium falciparum, the deadliest species of malaria-causing parasites. In phase 1, Optibrium’s Augmented Chemistry™ technologies, which incorporate Intellegens’ Alchemite™  deep learning platform, were used to build accurate predictive models for activity against this target. These were applied to guide the design efforts in phase 2. Combining StarDrop™ and Augmented Chemistry™ technologies, the team designed the only compound, out of four submitted by different organisations, for which activity was confirmed using in vitro tests, and the measured activity was in strong agreement with the predicted values.
Founded in 2012 by Professor Matthew Todd, Chair of Drug Discovery at University College London, the OSM consortium aims to find new medicines for the treatment of malaria, which is recognised by the World Health Organisation as one of the world's biggest killers. The latest results from the initiative can be found here.
Dr Benedict Irwin, Senior Scientist at Optibrium, said:
“Our latest work with the Open Source Malaria consortium is a testament to the power of Optibrium’s software. It demonstrates, in an open and transparent way, the impact this dynamic blend of computational chemistry and machine learning can have in supporting drug discovery scientists in tackling these serious diseases.”
Professor Matthew Todd, Founder of the OSM consortium, added:
“It’s great to see that the Optibrium/Intellegens’ strong modelling results from phase 1 could be complemented with generative methods and held up in in vitro testing. While the use of AI in drug discovery is still in its infancy and in many cases the potential of in silico designed compounds hasn’t yet been rigorously validated experimentally, this example can help pave the way and is a valuable contribution to our efforts. I hope to see more from the team in support of our quest to develop effective treatments for malaria.”
Dr Tom Whitehead, Head of Machine Learning at Intellegens (pictured), said:
“This result is a powerful validation of the benefit advanced deep learning methods such as Alchemite™ can bring to chemistry design and optimisation problems. We are looking forward to continuing to support OSM in their pursuit of new treatments for malaria.”
 Whitehead et al. J. Chem. Inf. Comput. Model. (2019) 59(3) pp. 1197-1204
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. Their mission is to help their clients accelerate innovation by using their unique deep learning solutions to extract valuable information from existing processes and data. The technique, created at the Cavendish Laboratory, is encapsulated in Intellegens’ first commercial product, Alchemite™ . The cutting-edge deep learning algorithms that Alchemite™ 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, Alchemite™ can unravel data problems that are not accessible to traditional deep learning approaches. Suitable for deployment across any kind of numeric dataset, Alchemite™ 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.