A library of the paired peptide/HLA multimers and artificial APCs allows for construction of a large database of class I-restricted peptides and cognate tumor-reactive TCR genes at an unprecedented scale.

A new, high-throughput in vivo MHC-I peptide minigene library platform shows that the naive immune system cannot eliminate cells presenting immunogenic antigens found at low frequencies within a growing tumor.

Tumor immunogenicity is quantified with a novel method, and the resulting tumor immunogenicity score is an effective tumor-inherent biomarker for prediction of response to immune checkpoint inhibitors.

Human NK cells engineered to express a PD-L1 chimeric antigen receptor can control murine and human tumors and reduce myeloid cells expressing high levels of PD-L1.

An immunopeptidomic-based pipeline provides a tool for the identification of cancer therapeutic targets and generation of cancer oncolytic vaccines.

Amino acid codon degeneracy indicates antigen-specific selection of the T cell receptors and is predictive of cancer immunotherapy outcomes.

Loss of membrane-bound ICAM-1, overexpression of soluble ICAM-1 or ILKAP activity in tumor cells induce resistance to antigen-specific cytotoxic T-cell-mediated killing independently from regulating antigen presentation, PD-L1 expression, IFN-γ, or TNF-α responsiveness.

Cluster analysis reveals convergent T cell clones reacting to tumor antigens, enabling intelligent optimization of cancer immunotherapy.

Theoretical modeling reveals how stochastic optimal cancer immune evasion via antigen downregulation gives rise to diverse post-escape cancer antigenic profiles in a manner dependent on the immune microenvironment.

Modeling the dynamics of immunological synapses informs pharmacodynamics and treatment resistance to bispecific T cell engagers.