Calibrated Abstention for Reliable TCR--pMHC Binding Prediction under Epitope Shift

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bekov, Arman, Bekzhanov, Timur, Sadykov, Bekzat
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918447191949312
author Bekov, Arman
Bekzhanov, Timur
Sadykov, Bekzat
author_facet Bekov, Arman
Bekzhanov, Timur
Sadykov, Bekzat
contents Predicting T-cell receptor (TCR)--peptide-MHC (pMHC) binding is central to vaccine design and T-cell therapy, yet deployed models frequently encounter epitopes unseen during training, causing silent overconfidence and unreliable prioritization. We address this by framing TCR--pMHC prediction as a \emph{selective prediction} problem: a calibrated model should either output a trustworthy confidence score or explicitly abstain. Concretely, we (1) introduce a dual-encoder architecture encoding both CDR3$α$/CDR3$β$ and peptide sequences via a pre-trained protein language model; (2) apply temperature scaling to correct systematic probability miscalibration; and (3) impose a conformal abstention rule that provides finite-sample coverage guarantees at a user-specified target error rate. Evaluated under three split strategies -- random, epitope-held-out, and distance-aware -- our method achieves AUROC 0.813 and ECE 0.043 under the challenging epitope-held-out protocol, reducing ECE by 69.7\% relative to an uncalibrated baseline. At 80\% coverage, the selective model further reduces error rate from 18.7\% to 10.9\%, demonstrating that calibrated abstention enables principled coverage-risk trade-offs aligned with practical screening budgets.
format Preprint
id arxiv_https___arxiv_org_abs_2604_13254
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Calibrated Abstention for Reliable TCR--pMHC Binding Prediction under Epitope Shift
Bekov, Arman
Bekzhanov, Timur
Sadykov, Bekzat
Graphics
Predicting T-cell receptor (TCR)--peptide-MHC (pMHC) binding is central to vaccine design and T-cell therapy, yet deployed models frequently encounter epitopes unseen during training, causing silent overconfidence and unreliable prioritization. We address this by framing TCR--pMHC prediction as a \emph{selective prediction} problem: a calibrated model should either output a trustworthy confidence score or explicitly abstain. Concretely, we (1) introduce a dual-encoder architecture encoding both CDR3$α$/CDR3$β$ and peptide sequences via a pre-trained protein language model; (2) apply temperature scaling to correct systematic probability miscalibration; and (3) impose a conformal abstention rule that provides finite-sample coverage guarantees at a user-specified target error rate. Evaluated under three split strategies -- random, epitope-held-out, and distance-aware -- our method achieves AUROC 0.813 and ECE 0.043 under the challenging epitope-held-out protocol, reducing ECE by 69.7\% relative to an uncalibrated baseline. At 80\% coverage, the selective model further reduces error rate from 18.7\% to 10.9\%, demonstrating that calibrated abstention enables principled coverage-risk trade-offs aligned with practical screening budgets.
title Calibrated Abstention for Reliable TCR--pMHC Binding Prediction under Epitope Shift
topic Graphics
url https://arxiv.org/abs/2604.13254