An AI-powered pipeline for personalized cancer vaccines

Ludwig Cancer Research scientists have developed a comprehensive, start-to-finish computational pipeline that integrates multiple molecular and genetic analyzes of tumors and specific T cell molecular targets and leverages artificial intelligence algorithms to use its results to design personalized cancer vaccines for patients.

The design, validation and comparative evaluation of this computing suite, NeoDisc, are detailed in the current issue of Natural biotechnology in a publication led by Florian Huber and Michal Bassani-Sternberg from the Lausanne branch of the Ludwig Institute for Cancer Research.

“NeoDisc provides unique insights into the immunobiology of tumors and the mechanisms by which they evade targeting by cytotoxic T cells of the immune system,” said Bassani-Sternberg.

“This information is invaluable for the design of personalized immunotherapies, and the analytical and computational pipeline at the heart of NeoDisc is already in use here in Lausanne for clinical trials of personalized cancer vaccines and adoptive cell therapies.”

Many types of cancer harbor multiple random mutations that should make them more visible to the immune system. Such mutations generate aberrant proteins that cells, even cancer cells, are programmed to cut into small pieces – called peptides – and “present” as antigens to provoke an attack by patrolling T cells.

The great diversity of these “neoantigens” is one of the reasons why patients respond so variably to immunotherapies. On the other hand, neoantigens can be harnessed to develop vaccines and other types of immunotherapies tailored to uniquely target each patient’s tumors. Personalized treatments like this are now being developed by researchers around the world.

Such efforts are technically difficult, because not all neoantigens are recognized by a given patient’s T cells, and even many fail to elicit a sufficiently potent attack on T cells. One approach to designing Personalized vaccines and cellular therapies therefore involve the identification of neoantigens most likely to provoke a vigorous attack on T cells.

This requires sophisticated, large-scale analyzes of the mutations that generate potential neoantigens, the molecular scaffolding (called HLA molecules) that presents them to T cells, and the molecular features that enable recognition by T cell receptors. Bassani- Sternberg is among the pioneers of this field, a high-tech marriage of large-scale biochemical and computational analysis known as “immunopeptidomics.”

The design of personalized immunotherapies is also facilitated by the genomic analysis of the tumor and blood cells that represent the patient’s healthy genome, the large-scale analysis of gene expression known as “transcriptomics” as well as as sensitive analysis of the called immunopeptidome with mass spectrometry.

However, until now, these powerful technologies have never been integrated into a single computational pipeline to predict which neoantigens identified in a patient’s tumors should be used as vaccines or otherwise exploited for personalized immunotherapies.

Beyond this, neoantigens are not the only type of antigen available for immunotherapeutic targeting. Cancer cells also erroneously express as proteins usually non-coding DNA fragments, genes normally expressed only during development, other aberrantly expressed gene products, and viral antigens, in the case of cancerous tumors. of viral origin, all of which can cause an immune attack.

“NeoDisc can detect all of these distinct types of tumor-specific antigens as well as neoantigens, applying machine learning and rule-based algorithms to prioritize those most likely to provoke a T cell response, then use this information to design a personalized cancer vaccine for the respective patient,” Huber said.

NeoDisc further classifies the potential antigens it detects and generates visualizations of cancer cell heterogeneity within tumors.

“Notably, NeoDisc can also detect potential defects in the antigen presentation machinery, alerting vaccine designers and clinicians to a key mechanism of immune evasion in tumors that can compromise the effectiveness of immunotherapy,” Bassani-Sternberg said. “This can help them select patients for clinical studies who may benefit from personalized immunotherapy, a capability that is also of great importance for optimizing patient care.”

The researchers further show in their study that NeoDisc provides a more precise selection of effective cancer antigens for vaccines and adoptive cell therapies than other computational tools currently used for this purpose.

To further improve NeoDisc’s accuracy, researchers will continue to feed it data obtained from a variety of tumors and integrate additional machine learning algorithms into the software suite to advance its training and improve its predictive accuracy.