Peter Doyle, BioSimulytics
AI-enhanced molecular modelling for smarter, faster and more cost-effective drug development
What problem are you solving and what is innovative about your approach?
Drug molecules are manufactured in their solid crystal state, but crystal structures are complicated by polymorphism, the ability of a compound to exist in more than one stable crystal structure.
A simple ice (H2O) molecule for example has 18 stable structures; drug molecules, which are vastly more complex compounds, can have thousands of possible stable structures and several viable ones for market use.
To make it even more complicated, a polymorph may change to a more thermodynamically stable form hours, weeks and even years later depending on conditions.
Different polymorphs can have different properties such as solubility or toxicity, which makes it so vital for pharma companies to fully understand the polymorphic landscape of their drug molecules and to have absolute certainty that the polymorph they bring through clinical trials and manufacturing is the one that stays the same in the marketplace.
In fact, precise definition of the polymorph is essential nowadays for regulatory compliance as well as patent protection.
BioSimulytics has developed unique software which only requires the basic 2D structure of a compound to accurately predict the detailed profiles of all its 3D polymorphic forms, ranked by the most stable, with levels of speed and data far beyond what is possible with current experimentation methods. The BioSimulytics solution uses a powerful combination of computational chemistry, quantum physics, high-performance computing and Artificial Intelligence (AI).
How is this idea commercially attractive?
The scope for AI to revolutionise drug development is immense. Typically it takes US$2.6bn and 10-14 years to bring a new drug to market with 1% chance of success. The scope for innovation is huge and the pressure to do things differently is increasing. New digital tools will be key to driving smarter, faster and more cost-effective R&D processes in the pharma industry. In particular, AI-based machine-learning and neural-network technologies are being increasingly deployed in the discovery, design and development of new drugs.
What do you hope to achieve by participating in Big Ideas?
BioSimulytics has the ambition to become a global leader in software tools for Digital R&D in the world of material sciences, exploiting its breakthrough technology for universal applications. The company will need strong like-minded partners and investors to help realize its ambition. Big Ideas provides a great opportunity to meet such people.