# Agretopicity

Source: https://onco.cc/terms/agretopicity/  
OnCo record `agretopicity` (Term). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

Agretopicity concerns how a protein fragment interacts with an immune display molecule. Comparing altered and normal fragments gives a separate predicted score; actual display and immune recognition need tests.

## Summary

Agretopicity concerns a peptide's ability to interact with a major histocompatibility complex (MHC) display molecule, called human leukocyte antigen (HLA) in people. It is distinct from the differential agretopic index (DAI), which compares predicted binding of the mutant peptide with its matching wild-type counterpart against the same molecule. Duan and colleagues named the numerical difference in NetMHC scores between mutated and unmutated sequences the differential agretopic index. They subtracted the wild-type NetMHC score from the mutant score. Ghorani and colleagues instead subtracted predicted mutant affinity from predicted wild-type affinity. Some groups use a ratio, so retain the units, formula and sign convention before comparing values across pipelines.

Two separate ideas sit behind the measure. First, a mutation at a position the HLA groove grips can create a new anchor residue and change binding, potentially changing presentation; Duan and colleagues found that neoepitopes mutated to create new anchor residues, together with greater predicted rigidity of the peptide and HLA complex, identified protective neoepitopes with relatively high accuracy in mouse tumours (the CMS5 and Meth A mouse tumours). Second, reasoning rather than a finding of the cited papers: a peptide that normal tissue never displayed well may not have had its reactive T cells removed by self-tolerance, so a high value is sometimes read as a proxy for an intact T-cell repertoire against that target.

A high value does not require strong absolute binding. In that mouse work the protective neoepitopes produced CD8-dependent immunity although their affinity for the mouse class I molecule H-2Kd was orders of magnitude weaker than the conventional 500 nanomolar threshold, which is part of the argument for ranking on the mutant-versus-wild-type comparison rather than on absolute predicted affinity alone.

Ghorani’s retrospective human analysis included 66 lung adenocarcinomas and 72 melanomas from The Cancer Genome Atlas, plus three immunotherapy-treated cohorts containing 131 melanoma patients in total and 31 with non-small-cell lung cancer. Mean DAI correlated with overall survival in three of five cohorts, with the reported effect for clonal rather than subclonal peptides. The authors reported evidence of superiority over nonsynonymous mutational and neoantigen burden in these analyses, not a universally superior clinical predictor. Mean DAI also correlated with immune-infiltration measures in melanoma (P = 0.027) and lung adenocarcinoma (P = 0.024), whereas mutational and neoantigen burden did not. The association with survival was not uniform: P = 0.068 in the TCGA melanoma cohort and P = 0.003 in one treated melanoma cohort. These population associations and the animal protection findings are different evidence tiers. Neither prospectively demonstrates that selecting a vaccine by DAI improves clinical outcomes.

The 2020 Tumor Neoantigen Selection Alliance study tested which peptide features predict immunogenicity: 608 predicted epitopes were assessed for T-cell binding in patient-matched samples, and a model integrating peptide features linked to presentation and recognition filtered out 98% of non-immunogenic peptides at a precision above 0.70, then held in an independent cohort of 310 epitopes (Wells and colleagues). Its abstract does not say which features entered the model, so this record does not claim that the study supported or rejected agretopicity.

## Fields

- Kind: Term
- Last checked: 2026-11-11
- Also known as: differential agretopic index; differential agretopicity index; DAI; differential binding affinity

## Notes

- Peptide binding to HLA and T-cell receptor recognition of the peptide-HLA complex are different interactions. The peptide's HLA-binding determinants can be called an agretope; its recognition by a receptor depends on the combined surface. This functional distinction does not prescribe two flat, non-overlapping faces for every complex. A binding prediction or mutant-versus-normal score is not a measurement of natural display, T-cell response, killing or clinical benefit.
- The differential agretopic index is not absolute binding strength: its reference is the peptide's own wild-type version. Sequence self-similarity asks a different question about resemblance to normal human sequences more broadly. The Duan and Ghorani studies concern class I and do not validate a class-II ranking rule. HLA is the human major histocompatibility complex, called MHC in the mouse experiments. Predicted binding differences also do not prove absence of self-tolerance.
- The score is computed from sequence and binding predictions for the alleles supplied by HLA typing. Retain the peptide windows, allele, predictor and version, both scores, units and subtraction or ratio rule. Sign conventions differ, so a larger number in one pipeline need not match another's scale. Computation adds no tissue assay, but depends on the sequence and typing inputs. On a design board it sits beside expression, clonality and measured presentation. Absolute affinity alone does not reveal how much the mutant differs from its wild-type counterpart. Neither comparison proves escape from tolerance, natural presentation or a useful T-cell response. Different predictors can give different estimates for the same peptide pair.

## Sources

- Duan et al., genomic and bioinformatic profiling of mutational neoepitopes reveals new rules to predict anticancer immunogenicity (J Exp Med 2014): https://pubmed.ncbi.nlm.nih.gov/25245761/
- Ghorani et al., differential binding affinity of mutated peptides for MHC class I is a predictor of survival in advanced lung cancer and melanoma (Ann Oncol 2018): https://pubmed.ncbi.nlm.nih.gov/29361136/
- Wells et al., key parameters of tumor epitope immunogenicity revealed through a consortium approach improve neoantigen prediction (Cell 2020): https://pubmed.ncbi.nlm.nih.gov/33038342/
- Wieczorek et al.: peptide binding and recognition of peptide-MHC complexes (2017): https://pubmed.ncbi.nlm.nih.gov/28367149/

## Connected records

- terms: [CD4 T-cell subsets](https://onco.cc/terms/cd4-t-cell-subsets/), [Interferon-gamma gene signature (T-cell-inflamed signature)](https://onco.cc/terms/interferon-gamma-signature/), [Neoantigen](https://onco.cc/terms/neoantigen/), [Neoantigen candidate selection](https://onco.cc/terms/neoantigen-candidate-selection/), [Tumour mutational burden (TMB)](https://onco.cc/terms/tmb/)
- technologies: [Personalised neoantigen (mRNA) vaccines](https://onco.cc/technologies/neoantigen-mrna-vaccine/)
- pathways: [The cancer-immunity cycle](https://onco.cc/pathways/cancer-immunity-cycle/)

---
JSON: https://onco.cc/api/v1/entities/agretopicity.json