Written by:
Senior Vice President, Biologics Engineering and Oncology Targeted Discovery, R&D, AstraZeneca
Executive Director, Biologics Design and Technology, Biologics Engineering, R&D, AstraZeneca
Senior Director, Emerging Technologies & High-Throughput Platforms, Biologics Engineering, R&D, AstraZeneca
How is AI helping scientists design the next generation of biologic medicines?
Developing a biologic medicine means finding the rare molecules, among vast numbers of possibilities, with the right set of properties to become a medicine. They must bind to the right disease target with the right affinity, at the right site, behave in the body as intended, and be suitable for manufacturing.
At AstraZeneca, AI is helping scientists assess and refine potential molecules earlier in biologics discovery, as we shift from traditional experimental workflows to AI-enabled drug discovery and automated approaches that enable scientists to spend more time focused on strategic drug discovery decisions. This is part of a broader approach to AI drug discovery, in which biology, laboratory automation, and computation work side by side.
Why is AI becoming more important in biologics discovery?
Advances in science and innovation mean that biologics discovery is becoming increasingly sophisticated, and so are the challenges involved in designing them. Scientists are now working across increasingly complex formats, including multispecific antibodies, antibody drug conjugates and cell therapies, where success depends on balancing many drug properties at once.
AI is well suited to this complexity. It can help interpret rich biological data, identify patterns across large datasets and predict molecular properties before a molecule is made. At AstraZeneca, this is helping scientists identify potential biologic molecules for development in a fraction of the time, enabling earlier decisions and more focused use of lab resources.
These capabilities support antibody engineering and computational drug design, helping scientists explore and refine biologic molecules with greater precision.
We are integrating AI across biologics discovery to help scientists navigate complexity earlier, make more informed design choices and focus resources on the candidates with the strongest potential.
What is de novo design?
One of the most promising frontiers in this field is de novo design, where generative and predictive AI are used to create entirely new antibody sequences from scratch and optimise them for important features such as binding, specificity and manufacturability. This is also known as de novo antibody design and is an emerging application of AI protein design.
AstraZeneca is seeing encouraging early results in this emerging area of science. These include AI-assisted library generation, optimisation of promising molecules and better prioritisation of candidates for development. Rather than relying only on naturally occurring molecules or incremental redesign, scientists can begin to explore novel options that may not be found through traditional approaches alone.
That matters because fit-for-purpose design earlier in discovery can improve the quality of the molecules that move forward. Over time, this could help improve the probability of success in development and support the goal of bringing more effective medicines to patients.
De novo design is a long-term scientific ambition, but its early value is already visible today in how AI helps us generate, refine and prioritise biologic candidates more effectively.
How do AI and automation work together in biologics discovery?
For AI to keep improving, it needs high-quality, standardised data. That is where automation becomes important. By connecting design, testing and analysis in a more integrated workflow, automation helps generate data more consistently and feeds each round of learning back into the next.
AstraZeneca is continuing to invest in automation across biologics discovery to support this approach, bringing together automation, standardised data pipelines and computational intelligence in more connected workflows – always with human oversight.
Automation helps us generate the consistent, high-quality data needed to strengthen AI models over time and accelerate learning across biologics discovery.
Together, AI, automation and human scientific expertise are helping reshape biologics discovery, with the aim of bringing innovative medicines to patients faster.
