Molecular informatics
Represent, organise and explore molecular information to support screening, comparison and knowledge discovery.
Chemistry AI & Computational Science
Bring intelligence closer to science.
Connecting chemical knowledge, scientific data and computational methods to make complex scientific exploration more informed and more useful.
01 / The opportunity
Scientific complexity is not a reason to stop exploring.
Chemistry produces rich, interconnected information—from molecular structures and reaction pathways to experimental observations and materials data.
AI and computational methods can help researchers organise that information, identify patterns, prioritise possibilities and ask more targeted questions. The value comes from combining these methods with domain knowledge, data quality and scientific validation.
02 / Areas of exploration
Computational tools for chemical questions.
Our Chemistry AI direction focuses on practical scientific use cases, with an emphasis on explainability, reproducibility and meaningful validation.
Property prediction
Explore data-driven approaches to estimate molecular or material properties, with appropriate attention to uncertainty and validation.
Materials informatics
Connect composition, processing, structure and performance data to support materials analysis and discovery workflows.
Reaction intelligence
Investigate computational approaches for reaction information, pathway exploration and chemical knowledge organisation.
Scientific data workflows
Structure, clean and analyse scientific datasets so they can be interpreted and reused more effectively.
Computational research tools
Build purpose-specific applications that connect scientific data, modelling workflows and researcher-facing interfaces.
03 / A disciplined approach
AI is a tool. Scientific judgement stays central.
We approach scientific AI as an evidence-led process rather than a shortcut to certainty.
01
Frame the question
Clarify the scientific objective, the available evidence and what a useful output would look like.
02
Understand the data
Review provenance, coverage, representation and limitations before selecting a method.
03
Build & evaluate
Develop the computational workflow and evaluate it against appropriate baselines and scientific criteria.
04
Interpret responsibly
Communicate results with limitations, uncertainty and the need for expert validation made clear.
Chemistry × computation