In a groundbreaking collaboration announced this week at IBM Research Zurich, IBM and Boehringer Ingelheim are set to revolutionize the field of therapeutic antibody development by harnessing the power of generative AI. The partnership aims to leverage IBM's advanced AI technologies to accelerate the discovery of novel candidate antibodies, ultimately improving patient outcomes.
The core of this collaboration involves Boehringer Ingelheim utilizing IBM's AI models, which are trained on billions of proteins in a self-supervised environment. The objective is to expedite the health discovery process by generating hypotheses faster and more effectively, with the ultimate goal of reducing antibody discovery timelines from the conventional three years to an impressive one year.
Boehringer Ingelheim, a leading research-driven biopharmaceutical company dedicated to innovation in areas of high unmet medical need, will employ IBM's pre-trained AI model, fine-tuned on additional data from the company. This innovative approach using foundation models for antibody discovery is poised to transform the landscape of healthcare, potentially benefitting patients worldwide while streamlining the work for scientists.
The discovery and development of therapeutic antibodies, crucial in treating diseases like cancer, autoimmune, and infectious diseases, have historically been a complex and time-consuming process. Despite significant technological advances, the traditional timeframe of three years for antibody discovery, often followed by repetition, has remained a challenge. The introduction of AI into this process holds the promise of significantly reducing discovery timelines, offering a potential paradigm shift.
Alessandro Curioni, Vice President Accelerated Discovery at IBM Research, expressed excitement about extending the impact of generative AI models beyond language and highlighted the potential for IBM's multimodal foundation model technologies to accelerate the creation of new therapeutics at Boehringer Ingelheim.
The collaboration extends beyond AI modeling to in-silico methods, combining molecular dynamics with AI to accelerate discovery processes. Boehringer Ingelheim, in conjunction with IBM researchers, aims to use in-silico methods to generate new human antibody sequences based on disease-relevant targets, molecular profiles, and success criteria for therapeutically relevant antibody molecules. These sequences will then be used to train foundation models, creating unique datasets.
The collaborative efforts will also involve laboratory experiments to validate and improve in-silico methods through feedback loops, further enhancing the effectiveness of the AI-driven discovery process. This collaborative approach aligns with Boehringer Ingelheim's vision to build a leading digital ecosystem, facilitating the acceleration of drug discovery and development and opening new opportunities to transform patients' lives.
Andrew Nixon, Global Head of Biotherapeutics Discovery at Boehringer Ingelheim, expressed confidence in the collaboration, stating that the partnership with IBM scientists will lead to the development of an unprecedented platform for accelerated antibody discovery, ultimately enabling Boehringer Ingelheim to deliver new treatments for patients with high unmet needs.
This collaboration not only underscores Boehringer Ingelheim's commitment to innovation but also highlights IBM's dedication to utilizing generative AI and foundation models to expedite discoveries within healthcare contexts. IBM researchers anticipate rapid progress, with model fine-tuning set to occur at an accelerated pace, leading to the testing of the first set of models within a couple of months. The collaborative effort stands as a testament to the transformative potential of combining cutting-edge technology with pharmaceutical expertise to reshape the landscape of therapeutic antibody development.
