AI-driven neoantigen identification: a comprehensive review from somatic variant calling to T cell recognition.
This review details AI workflows that combine transcriptomics and immunopeptidomics to nominate candidate neoepitopes, moving beyond tailored experimental methods. It emphasizes the critical challenge of predicting which tumor mutations result in T cell recognition via class I and II peptide-binding grooves. The piece serves as a technical baseline for understanding the computational layers required to reduce false positives in personalized vaccine design.