SwiftRepertoire: Few-Shot Immune-Signature Synthesis via Dynamic Kernel Codes

2026-02-01

AI Summary

T cell receptor (TCR) repertoire analysis holds great promise for disease detection, yet its application is hindered by sparse labels, cohort heterogeneity, and the high computational cost of full-model fine-tuning. To address these challenges, this work proposes a plug-and-play few-shot adaptation method that avoids end-to-end fine-tuning. By freezing a pretrained encoder and leveraging task descriptors to guide dynamic kernel codes, the approach synthesizes lightweight task-specific adapters from a learned prototype dictionary. It further integrates repertoire-level probes with embedding statistics to enhance representation. This framework enables efficient and interpretable immune modeling with minimal labeled samples, preserving the interpretability of sequence motifs while remaining suitable for clinical and research settings where labels are scarce and computational resources are limited.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Computer Vision: Representation Learning for Vision

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Repertoire-level analysis of T cell receptors offers a biologically grounded signal for disease detection and immune monitoring, yet practical deployment is impeded by label sparsity, cohort heterogeneity, and the computational burden of adapting large encoders to new tasks. We introduce a framework that synthesizes compact task-specific parameterizations from a learned dictionary of prototypes conditioned on lightweight task descriptors derived from repertoire probes and pooled embedding statistics. This synthesis produces small adapter modules applied to a frozen pretrained backbone, enabling immediate adaptation to novel tasks with only a handful of support examples and without full model fine-tuning. The architecture preserves interpretability through motif-aware probes and a calibrated motif discovery pipeline that links predictive decisions to sequence-level signals. Together, these components yield a practical, sample-efficient, and interpretable pathway for translating repertoire-informed models into diverse clinical and research settings where labeled data are scarce and computational resources are constrained.

Problem

• label sparsity
• cohort heterogeneity
• computational burden
• few-shot adaptation
• immune repertoire analysis

Innovation

• few-shot learning
• immune repertoire
• dynamic kernel codes
• adapter modules
• interpretable AI