Artificial intelligence and neoantigens: paving the path for precision cancer immunotherapy

Abstract

Cancer immunotherapy has witnessed rapid advancement in recent years, with a particular focus on neoantigens as promising targets for personalized treatments. The convergence of immunogenomics, bioinformatics, and artificial intelligence (AI) has propelled the development of innovative neoantigen discovery tools and pipelines. These tools have revolutionized our ability to identify tumor-specific antigens, providing the foundation for precision cancer immunotherapy. AI-driven algorithms can process extensive amounts of data, identify patterns, and make predictions that were once challenging to achieve. However, the integration of AI comes with its own set of challenges, leaving space for further research. With particular focus on the computational approaches, in this article we have explored the current landscape of neoantigen prediction, the fundamental concepts behind, the challenges and their potential solutions providing a comprehensive overview of this rapidly evolving field.

Keywords: artificial intelligence; cancer immunotherapy; immunopeptidomics; neoantigen prediction; precision medicine.

Figure 1

Schematic overview of AI algorithm training on public databases. A group of subjects, specific for the condition of interest is chosen for the experimental procedures. After completing the experimental pipelines, the generated data is stored in a public database. AI algorithms can then be trained on these datasets.

Figure 2

Steps of neoantigen selection from patient data. A set of diagnostic procedures are completed on patient derived samples. Ideally all of the above-mentioned patient data (WES, WGS, HLA typing, RNA-seq) are available before proceeding. After a candidate peptide selection is generated from the patient data, the AI model of preference is applied. The AI model will compute a ranked peptide list from the candidate peptides. Careful design of personalized vaccine is available, based on the peptide rankings.

Figure 3

NGS-based HLA genotyping. Sequence data generated by sequencing technologies is mapped against the reference allele repository (IPD-IMGT). Corresponding to the HLA genotyping algorithm used either the raw reads or assembled contigs are aligned.

Figure 4

Challenges and potential solutions to promote widespread clinical use of AI applications for neoantigens discovery. We distinguish challenges that must be addressed for successful AI integration into clinical praxis as related to data, models, AI architecture and technical integration. For each group of challenges we list various algorithmic, experimental and organizational approaches carrying the potential to overcome the respective challenges.