Unbiased Neoepitope Identification via Bivariate IC50 Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying tumor rejection mediating neoepitopes (TRMNs) are inefficient, as they rely on MHC-peptide interactions derived from viral epitopes, which do not accurately predict TRMNs, leading to ineffective cancer vaccines with no correlation with overall survival or tumor rejection.
Innovation Solution
An unbiased method involving comparison of cancer cell and reference exome sequences to identify single nucleotide variants, calculating IC50 values for MHC binding, and selecting neoepitopes in a specific bivariate scatter plot space to identify TRMNs, which are then used to create a peptide or nucleic acid population for cancer treatment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If MHC-peptide interaction prediction algorithms are used to identify neoepitopes, then the identification process is simplified and follows established viral epitope rules, but the accuracy of identifying true tumor rejection mediating neoepitopes deteriorates
Solution Approach 1:
The patent changes the selection parameters from MHC binding affinity (IC50 < 50 nM) to a dual-criterion system requiring both moderate MHC binding (IC50 > 501 nM for mutant peptide) and significant differential binding compared to wild-type (IC50 ratio > 10). This parameter transformation resolves the contradiction by abandoning the viral epitope-derived threshold in favor of tumor-specific selection criteria that actually predict tumor rejection.
Solution Approach 2:
The patent inverts the conventional approach by selecting neoepitopes with LOWER MHC binding affinity rather than higher affinity. Traditional methods select for strong binders (low IC50), but this patent identifies that true TRMNs have moderate binding (IC50 > 501 nM) yet show significant differential binding compared to wild-type sequences. This inversion resolves the contradiction by rejecting the viral epitope rule that high affinity equals immunogenicity.
2Reliability
If high affinity MHC-binding neoepitopes are selected for cancer vaccines, then the vaccine composition is optimized for strong T cell recognition, but the clinical outcome deteriorates with no correlation to overall survival or tumor rejection
Solution Approach 1:
The patent changes the efficacy prediction parameters by introducing the IC50 ratio (mutant/wild-type binding affinity) as the key predictor rather than absolute binding strength. The selection criteria require IC50 > 501 nM for the mutant peptide and IC50 ratio > 10, which identifies neoepitopes that are differentially recognized compared to self. This resolves the contradiction by showing that differential recognition, not absolute affinity, predicts clinical efficacy.
Solution Approach 2:
The patent introduces the IC50 ratio as an intermediary parameter that mediates between MHC binding affinity and clinical efficacy. Rather than directly selecting for high affinity binders, the ratio serves as a filter that identifies neoepitopes with sufficient differential binding to elicit anti-tumor responses while maintaining adequate MHC presentation. This intermediary resolves the contradiction by decoupling absolute affinity from efficacy prediction.
3Quantity of substance
If CD8+ T cell response is used as the primary selection criterion for neoepitopes, then the vaccine targets strong immunogenic antigens, but the ability to identify true tumor rejection mediating neoepitopes deteriorates
Solution Approach 1:
The patent segments the selection process into multiple independent criteria: (1) MHC binding affinity (IC50 > 501 nM), (2) differential binding ratio (IC50 ratio > 10), and (3) absence of CD8+ T cell response in vitro. This segmentation resolves the contradiction by showing that true TRMNs can be identified by combining these criteria, even though individual CD8+ responsive neoepitopes fail to predict tumor rejection. The segmentation allows identification of a broader set of candidates that meet all criteria.
Solution Approach 2:
The patent inverts the CD8+ T cell response criterion by selecting for neoepitopes that do NOT elicit strong CD8+ responses in vitro. This counterintuitive approach identifies TRMNs that may require in vivo context or different mechanisms for tumor rejection. This inversion resolves the contradiction by rejecting the assumption that strong in vitro CD8+ responses predict in vivo tumor rejection.
Data Source
AI summary
Described herein is an unbiased method of identifying tumor rejection mediating neoepitopes (TRMNs). Putative neoepitopes from a cancer cell exome sequence from a cancer patient are putative neoepitopes are unbiased by MHC binding and/or CD8T* reactivity. By plotting the putative neoepitope IC50s on one axis, and the non-mutated amino acid sequence IC50s on a perpendicular axis to provide a bivariate scatter plot, novel TRMNs are identified TRMNs the neoepitopes in the bivariate scatter plot which are in the space greater than 501 nM on the x-axis and greater than 501 nM on the y-axis. Peptides and nucleic acids for expressing peptides including the TRMNs are also described.