Credit: Nastco/Getty Images
Aiming to address gaps in publicly available pipelines to predict clinically relevant neoantigen targets, researchers at the Ludwig Institute for Cancer Research are conducting both multiple molecular and genetic analyzes of tumors. We’ve developed a fully AI-enabled computational pipeline from start to finish that integrates. Mass spectrometry data for personalized cancer vaccine development.
Details of the new computational tool, called NeoDisc, were recently published in the journal Nature Biotechnology.
“NeoDisc provides unique insights into the immunobiology of tumors and the mechanisms by which tumors evade targeting by the immune system’s cytotoxic T cells,” said Dr. said lead author Dr. Michal Bassani-Sternberg. For cancer research. “These insights will be invaluable for the design of personalized immunotherapies, and NeoDisc’s core analytical and computational pipeline is already in clinical use for personalized cancer vaccines and adoptive cell therapies here in Lausanne. It is used for testing.”
Many types of cancer have multiple random mutations, which should make them easier for the immune system to recognize. These mutations produce abnormal proteins that cells are programmed to cut, even cancerous ones, into short fragments known as peptides, which are then ‘presented’ as antigens to T cells. provokes attacks by
These neoantigens are used in the development of today’s immunotherapies and vaccines to treat cancerous tumors, but the diversity of these peptides makes it difficult for immunotherapeutic approaches to treat most cancer patients. This is one of the main reasons why it is ineffective.
Because not all neoantigens are recognized by all patient T cells, and even if they are, they may not elicit a T cell response strong enough to effectively attack the cancerous tumor. Developing more personalized treatment approaches will be a major challenge.
To design personalized immunotherapies, researchers need to identify cancer mutations that generate potential neoantigens, HLA molecules that present them to T cells, and molecular signatures that allow them to be recognized by T cell receptors. A large-scale and detailed analysis must be conducted. The potential for personalized immunotherapy is informed by analysis of tumors and blood cells representing the patient’s healthy genome, transcriptomics, and analysis of the so-called immune peptide dome performed by mass spectrometry.
NeoDisc integrates genomics, transcriptomics, and immunopeptidomics into a single pipeline and uses AI to predict which neoantigens identified in a given patient should be utilized to create personalized cancer vaccines. One of the first technologies. NeoDisc can detect different types of tumor-specific antigens, including neoantigens, and apply machine learning and rule-based algorithms to prioritize antigens most likely to elicit a T-cell response .
“NeoDisc detects all these different types of tumor-specific antigens, along with neoantigens, and applies machine learning and rule-based algorithms to prioritize and identify the antigens most likely to elicit a T-cell response. We can use the information to design personalized cancer vaccines for patients,” said lead author Florian Huber, a bioinformatician at the Ludwig Institute.
Additionally, NeoDisc ranks the potential antigens it detects and visualizes the heterogeneity of cancer cells within the tumor.
“In particular, NeoDisc can also detect potential defects in the antigen-presenting machinery, alerting vaccine designers and clinicians to important mechanisms of immune evasion in tumors that can compromise the efficacy of immunotherapy.” Bassani・Mr. Sternberg stated. “This will help select patients for clinical studies who are likely to benefit from personalized immunotherapy, which is also a critical feature in optimizing patient care. ”
This study demonstrated that NeoDisc can more accurately select effective cancer antigens for vaccines and adoptive cell therapies than other computational tools currently in use.
Although NeoDisc provides promising initial results, the Ludwig team acknowledges its limitations, stating that “the quality and depth of the input data can significantly impact the accuracy of antigen identification and prioritization.” says. Nevertheless, this computational framework is currently being used in Phase I clinical trials of personalized cancer vaccines and adoptive T-cell therapy in Switzerland, providing major advances in tailoring cancer treatments to individual patients. is bringing. However, the researchers emphasized that different regulations may apply for clinical practice in other regions.
The Ludwig researchers’ next steps are to continue feeding NeoDisc with data from different tumor types and to integrate and continue training additional machine learning algorithms to improve the accuracy of predictions.
