What Is a Neoantigen? A Practical Guide for Experimental Researchers

By Lociven · NeoantigenLab · July 2026

If you work in cancer immunology, you have heard the word neoantigen more times than you can count. It appears in grant proposals, immunotherapy papers, and tumor board discussions. But the explanations are usually either too shallow ("mutated peptides the immune system can recognize") or buried in computational methods papers that assume you already know what a VCF file is.

This post is for the wet-lab side of the room — the researchers who understand tumor biology and immunology but want a clear picture of what neoantigens are, where they come from, and why the field cares so much about them.

Figure 1. Most somatic mutations do not become immunogenic neoantigens. Each step — coding region, MHC binding, processing, T cell recognition — acts as a filter.

The core idea

A neoantigen is a peptide fragment, derived from a tumor-specific somatic mutation, that can be presented on the surface of a cancer cell via MHC molecules and recognized by T cells.

Three things make this definition worth unpacking:

Tumor-specific somatic mutation. Normal cells carry the germline genome. Tumor cells accumulate mutations — substitutions, indels, frameshifts — that produce altered proteins not present anywhere in healthy tissue. These are somatic mutations. When the proteasome degrades these altered proteins, it can generate peptide fragments that differ from anything in the normal proteome.
MHC presentation. For a T cell to recognize a mutated peptide, that peptide has to be loaded onto an MHC molecule (HLA in humans) and displayed on the cell surface. Not every mutated peptide makes it to the surface — only those with the right binding affinity for the patient's specific HLA alleles.
T cell recognition. Even if a peptide is presented, it has to be recognized by a T cell receptor. Peptides derived from germline sequences are purged during thymic selection. Neoantigens, because they are truly foreign — absent from normal tissue — can escape central tolerance and activate T cells.

This is why neoantigens matter: they are the closest thing to a truly tumor-specific target the immune system has.

Where neoantigens come from

The most common source is single nucleotide variants (SNVs) — point mutations that change one amino acid in a protein. If the resulting peptide binds an HLA allele, it becomes a candidate neoantigen.

Other sources are less common but increasingly studied:

Insertions and deletions (indels) — particularly frameshift indels, which alter every downstream amino acid and can generate long stretches of novel sequence. These are often highly immunogenic because the resulting peptides look nothing like normal human proteins.
Gene fusions — chimeric proteins from chromosomal translocations create junction peptides that are absent from the normal proteome.
Splice variants — tumor-specific alternative splicing can expose cryptic exons.
Post-translational modifications — though harder to predict computationally, certain aberrant modifications can also generate novel epitopes.

In practice, most neoantigen identification pipelines focus on SNVs and indels because these are reliably detected from whole-exome sequencing (WES) data.

Not every mutation becomes a neoantigen

This is the part that surprises many researchers when they first see neoantigen prediction results.

A tumor with 200 somatic mutations might yield only 5–20 predicted neoantigens that are worth pursuing experimentally. The attrition happens at several steps:

The mutation has to be in a coding region. Synonymous (silent) mutations produce the same amino acid and generate no novel peptide.
The altered peptide has to bind the patient's HLA alleles. HLA is the most polymorphic locus in the human genome — each patient carries a unique combination of alleles with distinct peptide-binding preferences. A peptide that binds HLA-A*02:01 tightly might not bind HLA-A*03:01 at all.
The peptide has to be processed and presented. Binding affinity to HLA in silico is necessary but not sufficient. The peptide also has to survive proteasomal cleavage and TAP transport.
A cognate T cell has to exist in the patient's repertoire. Even a well-presented neoantigen is irrelevant if no T cell with the right TCR is available.

This funnel is why computational prediction alone is not enough — and why experimental validation (ELISpot, tetramer staining, T cell expansion assays) remains essential.

Shared vs. personal neoantigens

Most neoantigens are personal — unique to an individual patient because they arise from private somatic mutations. This is what makes neoantigen-based vaccines logistically challenging: each vaccine has to be custom-manufactured for each patient.

A small number of mutations recur across patients. The KRAS G12D/G12V mutations found in pancreatic and colorectal cancers are the most prominent examples. These shared neoantigens can in principle be targeted with off-the-shelf products, which is why they attract disproportionate clinical interest.

The Moderna/Merck mRNA-4157 trial — which showed a significant improvement in recurrence-free survival in melanoma when combined with pembrolizumab — used a personalized approach: up to 34 mutation-derived neoantigens per patient, encoded in a single mRNA construct. That trial changed how the field thinks about the commercial viability of personalized vaccines.

Why this matters beyond vaccines

Neoantigen research is not only about vaccine development. The same biology underlies:

Checkpoint inhibitor response prediction. Tumors with high TMB (tumor mutational burden) tend to respond better to anti-PD-1/PD-L1 therapy — partly because more mutations mean more neoantigens, meaning more pre-existing T cell responses that can be unleashed by checkpoint blockade.
TIL therapy. Tumor-infiltrating lymphocyte products (like Iovance's lifileucel) are enriched for T cells that recognize neoantigens. The clonality of neoantigen-reactive TILs correlates with clinical outcomes.
Adoptive T cell therapy. Identifying the neoantigens a patient's T cells already recognize allows engineers to select or engineer TCRs targeting those epitopes.

Understanding neoantigen biology is, in other words, increasingly prerequisite knowledge for anyone working in solid tumor immunotherapy — regardless of whether their lab focuses on vaccines specifically.

What comes next

The next post in this series covers tumor mutational burden (TMB) — how it is measured, what it does and does not tell you, and how to interpret TMB reports from clinical sequencing panels.

After that, we will walk through the full neoantigen identification pipeline: from WES data to a ranked list of candidate peptides, with each computational step explained in terms of what it is actually doing biologically.

From the SpatiaBio sister blog

If you work with spatial transcriptomics data

The SpatiaBio blog covers Squidpy-based spatial analysis with runnable notebooks — neighborhood enrichment, co-occurrence, ligand-receptor interactions, and the complete pipeline.

Visit SpatiaBio →

Tags: neoantigen, tumor immunology, cancer immunotherapy, MHC, HLA, somatic mutation, neoantigen vaccine, TMB, TIL

NeoantigenLab