Although widely used, cheminformatics is not confined just to pharmaceutical research – it is an irreplaceable tool for innovation in materials science. Innovative materials are crucial for tackling a range of challenges, from developing clean energy solutions to advancing medical devices. Historically, materials discovery was compared to ‘finding a needle in the haystack’ – it is notoriously difficult to find the balance of optimum properties, which often happened through trial-and-error approaches which took many years of research. With the rise of artificial intelligence (AI) and machine learning (ML), a new paradigm starts to appear: a shift from trial-and-error to data-driven innovation, where cheminformatics plays a central role in scaling materials discovery across industries.
Recognising the need for acceleration of the materials discovery process, initiatives such as The Materials Genome Initiative (MGI in the US) were created. MGI’s principles closely relate to cheminformatics, highlighting the need for data standardisations (through FAIR data practices) to be implemented in machine learning and computational modelling. One of their initiatives is ‘The Materials Project’, which provides open access to computed data on existing and predicted materials, driving the computational discovery of novel materials.
By combining molecular modelling, machine learning, and data mining, companies like Hitachi High Tech, Resonac and Schrödinger are developing platforms that accelerate the design of a wide range of materials – while also capitalising on the opportunity to patent their cheminformatics inventions.
Startups are also supercharging materials discovery. For example, Kebotix integrates multiple technologies including cheminformatics and robotics to create a closed-loop self-driving lab for automated design and generation of safer chemicals and high-performance materials. Another startup – Orbital Materials – uses a proprietary generative AI models and large-scale simulations to generate new sustainable materials. In partnership with AWS, they are aiming to develop advanced materials and climate technologies to tackle the data centre decarbonization and efficiency. MatNex, based in London, UK, employ proprietary AI software combined with rapid automated processes to accelerate design, development and testing of materials tuned to solve specific real-world problems.
The Big Tech companies have also shown a significant interest in innovating in the materials space. In 2023, Google released their Graph Networks for Materials Exploration (GNoME) model – an advanced graph neural network AI tool, designed to address the fundamental challenge of predicting the stability of new materials. In their paper published in Nature, they disclose the discovery of over 2.2 million previously unknown inorganic structures, which can be equated to 800 years’ worth of conventional materials discovery.
Earlier this year, Microsoft took it a step further by developing a tool which is able to produce new materials with specific desired properties, such as strength, conductivity, or magnetism. MatterGen uses a new type of architecture too – a diffusion model, which starts with random molecular structures and gradually ‘cleans up’ their 3D geometry to form a realistic structure. It can be paired with its companion tool – MatterSim, a deep-learning model which allows to computationally predict a material’s properties before making it in the lab. Importantly, MatterGen is open source, in the spirit of the collaborative vision driving modern materials research.
The shift from ‘trial-and-error’ to ‘data-driven’ discovery in materials science is a significant catalyst to development of new materials in a much reduced timeframe. Alongside that is a corresponding surge in development of IP.
As companies and startups harness AI-powered tools to design, model, and validate novel and protectable materials, it will be vital for them to also protect key underlying innovations at a foundational level.
For innovators and their advisors, this is an exciting time to shape the future of materials through both scientific ingenuity and strategic IP management.
This collaborative article was led by Anna Kukushkina, who interned in our Chemistry team last year.
Matthew is a Partner and Patent Attorney at Mewburn Ellis. Working primarily in the chemical and materials science fields, he has significant experience of the intricacies of the EPO. Matthew advises and assists clients with all stages of drafting, prosecution, opposition and appeal before the EPO. Many of his clients are Japanese and Chinese businesses that are seeking European patent protection. These include multinational corporations in the fields of high-performance ceramics and carbon fibre technologies, as well as pharmaceutical and cosmetic companies. Matthew also works with several research institutions and university technology transfer departments across Europe.
Email: matthew.smith@mewburn.com
