Astex Pharmaceuticals is a world leader in innovative drug discovery and development. The company has successfully applied its proprietary Fragment-Based Drug Discovery platform to generate multiple new drug candidates that are progressing in clinical development. Successful collaborations have led to three launched oncology drugs (Kisqali® partnered with Novartis, Balversa® partnered with Janssen and Truqap™ partnered with AstraZeneca). Astex continues to grow and focuses on Neurological Disorders and Oncology.

PROFILE AND SKILLS:

• PhD or equivalent experience in a technical discipline (e.g., computer science, chemistry, physics, engineering).
• Strong background in machine learning, including experience with deep learning and/or generative models.
• Proficiency with modern ML frameworks (e.g., PyTorch, TensorFlow, or JAX).
• Strong coding skills (e.g., Python, C++) and a collaborative mindset.
• Familiarity with protein structure modelling, co-folding, or related structure-based methods, along with a working knowledge of organic chemistry.

THE ROLE

We are seeking someone to drive the development of innovative machine learning methods for structure-based drug discovery. In this role, you will contribute to advancing computational tools that impact the design and optimisation of novel therapeutics.
You will work with one of the largest and most comprehensive internal structural datasets in the industry, our extensive collection of proprietary protein-ligand complexes—developing and applying cutting-edge AI techniques, including co-folding, to unlock new insights and capabilities in molecular modelling and virtual screening. This is a collaborative opportunity to push the boundaries of ML within drug discovery, working alongside multidisciplinary teams across computational chemistry, structural biology, and data science.

RESPONSIBILITIES:

• Design and implement ML models for structure-based design, including protein-ligand interaction modelling and co-folding applications.
• Develop and extend AI approaches that integrate structural and chemical data to improve virtual screening and molecular design workflows.
• Leverage Astex’s world-leading proprietary structural datasets to train, benchmark, and validate new algorithms.
• Collaborate closely with cross-functional teams to ensure effective translation of research into production-ready solutions.