IBM has one new collaboration with global research company Boehringer Ingelheim to use its AI base models to aid in the discovery of new antibody treatments.
Antibody treatments are the dominant weapon in the fight against a wide range of serious diseases and cause far fewer side effects compared to more traditional treatments.
It is hoped that by developing a fundamental AI model for antibody research, the production of life-saving treatments can become scalable and more efficient.
Generative AI for antibody discovery
Speaking at the launch event at IBM's research lab in Zurich, Alessandro Curioni, IBM Fellow, Vice President Europe and Africa and Director IBM Research Zurich, highlighted the potential benefits of fundamental models beyond AI and quantum, stating that “Foundation models can are used to address problems and data that lie outside the language domain.”
One of the major challenges hindering the development of new antibody treatments is the laboratory-based nature of antibody research. The successful development of a new antibody treatment requires controlled replication, meaning new treatments cannot be produced quickly enough, slowing the number of potential discoveries.
Therefore, IBM and Boehringer Ingelheim hope to scale up the development of new antibodies by using generative AI to produce molecules that have not been explored before, removing non-viable candidates and allowing researchers to focus on the promising candidates, increasing efficiency is significantly improved.
Furthermore, generative AI can help eliminate unconscious biases in the development of new antibodies by applying each candidate to a wide range of potential applications beyond an individual's area of expertise, significantly increasing not only the number of discoveries but also the number of use cases improved. for every discovery.