Friday, July 28, 2023
NEC Laboratories Europe and NEC Laboratories America have jointly developed an innovative artificial intelligence (AI) model called Attentive Variational Information Bottleneck (AVIB), which marks a significant step forward in the development of therapeutic cancer vaccines. AVIB is built upon previous research findings to predict the binding interactions between different proteins and molecules, a critical aspect of the human immune system's ability to recognize and combat cancer.
One of AVIB's primary applications is enabling biotechnologists to predict the binding between T-cell receptors and antigens presented on the surface of cancer cells. This predictive capability is vital for identifying neoantigens derived from cancer mutations, which can be used to create personalized therapeutic cancer vaccines. These vaccines teach the patient's immune system to recognize and target cancer cells for destruction. However, not all neoantigens that appear on the cancer cell's surface can efficiently bind to T-cell receptors and trigger the necessary immune response.
Dr. Pierre Machart, Senior Research Scientist at NEC Laboratories Europe, emphasizes the significance of this development, explaining that while many neoantigens from cancer mutations have the potential to elicit an immune response and be used in immunotherapy cancer vaccine design, identifying the most effective ones has been extremely challenging.
The traditional approach to developing immunotherapies involved laboriously testing which neoantigens are presented on the surface of a patient's cancer cells and if they could be recognized by the patient's T-cells. However, this method was time-consuming. With the use of machine learning models, this two-step process has been automated to some extent, but there have been limitations in predicting accurately which neoantigens would be recognized by T-cell receptors due to the scarcity of publicly available training data and the complexity of the molecules involved.
AVIB addresses this limitation and takes a significant step towards predicting the likelihood of T-cell receptors recognizing neoantigens on the surface of cancer cells. As a result, biotechnologists can more effectively rank the most promising vaccine elements based on the presentation of neoantigens and their binding potential with T-cells. Dr. Martin Renqiang Min, Department Head of Machine Learning at NEC Laboratories America, underscores that AVIB can significantly improve the efficacy of immunotherapy by enhancing the ranking of neoantigens within the therapeutic vaccine formula.
Beyond vaccine development, AVIB's predictive capability opens up new possibilities for T-cell therapy, a groundbreaking approach to immunotherapy. This involves engineering T-cells with specific receptors to bind directly to neoantigens on a patient's cancer cells, triggering their destruction. This advancement holds great promise for furthering the effectiveness of cancer immunotherapy and personalized treatments.