The validated drug discovery platform of Oxford Drug Design results from the distinctive integration of our two core competencies:
I. Comprehensive capabilities in aminoacyl-tRNA synthetases (aaRS)
II. Pioneering computational AI and machine learning methods
Each is distinctive in its own right, yet the seamless in-house combination of the two complementary strengths produces an even deeper and more differentiated capability for successful therapeutic discovery.
I. Deep Expertise in aminoacyl-tRNA Sythetases
Aminoacyl-tRNA synthetase enzymes (aaRS) play a key role in protein synthesis within all living organisms.
They also have key biological functions beyond protein synthesis, known as non-canonical functions, which include other essential processes like signalling and regulation.
Our comprehensive expertise in this area is a distinctive core competence of Oxford Drug Design, comprising:
- In-depth knowledge of aaRS structural biology, including proprietary X-ray crystallographic data
- Proprietary library of designed, novel, patent-protected chemotypes showing potent inhibitory activity
- Extensive in vivo and in vitro characterization
- Computational models (see below), focused specifically on aaRS
- Efficient networks of chemistry and biology with aaRS expertise
In both of our platform validation programs we have conclusively proven our design strategy by developing multiple chemotypes vs tRNA synthetases with strong in vivo activity.
Combined with our ML computational methods also trained symbiotically on the aaRS enzyme family, this deep understanding enables rapid and effective development of new therapeutics against multiple diseases with unmet need.
We are focusing first on cancer. Yet this expertise enables efficient expansion into other major therapeutic areas where aaRS enzymes also mediate disease.
II. Core Competence In Pioneering AI Drug Discovery Methods
AI and machine learning are transforming drug discovery.
Oxford Drug Design’s multi-level generative AI capability creates superior models to optimise molecular properties and thus obtain candidate drugs faster and more successfully.
We have developed completely new ways of representing the shape and properties of molecules that allow the efficient, accurate exploration of hundreds of millions of possible drug candidates. These derive from the latest developments in computational science and mathematics.
Our proprietary AI methods are powerful and diverse. They incorporate internal as well as public domain data which result in faster and more robust drug development progression using:
- Interpretable machine-learning technologies
- Pattern recognition, neural networks and
- Multi-parameter optimization
This differentiated technology capability builds on more than 10 years of foundational work in cheminformatics, bioinformatics and statistics, thus generating tried-and-tested methods that work in the practical domain.
More details are presented in our Technology section.