Ready for your all-in-one single cell sequencing solution?
Contact us

Predicting Epitope Specificity in Single Cell Immune Repertoire Data

8’
Immunology Infectious diseases

Understanding the adaptive immune response goes far beyond simply counting T-cell clones. To truly decode T-cell biology, you need to know not just which T-cells are expanding, but what they are responding to.

While identifying clones provides a snapshot of the immune repertoire, the epitope – and by extension the antigen- recognized by these clones remains the fundamental unit of functional significance. Whether the goal is to pinpoint the viral protein triggering an expansion or to identify the specific tumor antigen targeted by infiltrating lymphocytes, the ability to map a TCR to its cognate antigen is what transforms data into biological insight.

This functional understanding is critical across a variety of advanced applications:

  • T-cell therapy quality control: Characterizing tumor-infiltrating lymphocytes (TILs). Ensuring that the induced T-cells after treatment or in a product have the targeted specificity, and aren’t contaminated with bystanders.
  • Precision immunomonitoring: Tracking the evolution of a patient’s immune response in real-time during clinical trials to validate drug efficacy.
  • Target discovery in autoimmunity and oncology: Deconvoluting the complex TCR repertoires of patients to identify the “self-antigens” that elicit an immune response.

The Limits of Standard Profiling

Single cell immune profiling has revolutionized our ability to study T-cells by combining the T-cell transcriptome with their clonal TCR diversity. Platforms that generate full-length V(D)J sequences paired with whole-transcriptome data provide an incredibly high-resolution view of the immune repertoire.

However, this leaves a critical bottleneck. Standard bioinformatics pipelines excel at clonotype assembly and diversity quantification, but they cannot inherently interpret the biological meaning of a CDR3 sequence. But that leaves the most important question unanswered: What antigens are these T-cell receptors actually recognizing?

From Tissue Sample to Functional Annotation

To solve this bottleneck, Singleron Biotechnologies and ImmuneWatch have combined their respective expertise into a single, integrated workflow. By pairing Singleron’s highly precise single cell sequencing platforms with ImmuneWatch’s state-of-the-art immunoinformatics pipelines, you gain a seamless path from raw sample to functional epitope annotation.

The result is a transformation of raw sequence data into actionable immunological insight. You can bypass the high costs and slow timelines of traditional antigen discovery, allowing you to finally link clonal expansion directly to its functional target.

The Integrated Workflow: From Tissue to Target

The combined analytical pipeline consists of three phases: single cell partitioning, upstream processing, and downstream machine-learning-driven TCR or epitope specificity prediction.

Phase 1: Single Cell Partitioning and Library Preparation (Singleron)

The workflow begins with high-quality sample preparation. Using Singleron’s one-day workflow you can get from tissue to final single cell whole transcriptome and full-length VDJ libraries within one workday.

Tissue Processing Solid tissues are dissociated into viable single cell suspensions using manual protocols or automated systems like the PythoNi™. Cryopreserved cells can also be used, offering flexibility for longitudinal studies or multi-site collaborations.

Cell Partitioning and mRNA Capture Cells are loaded onto the SCOPE-chip® microfluidic platform, where they’re partitioned into individual microwells alongside barcoded capture beads. These beads capture poly-adenylated mRNA from each cell, including the targeted immunoreceptor transcripts. This poly-A-based approach ensures robust transcriptome coverage while simultaneously capturing TCR and BCR sequences.

Library Generation with sCircle® Here’s where Singleron’s technology delivers a distinct advantage. The sCircle® Single Cell Full-Length Immunoreceptor Library Kit uses a proprietary cDNA circularization process that brings the V(D)J region closer to the sequencing primer. This enables full-length sequencing of the entire V(D)J region, including the complete CDR3, as well as V, D, J, and C gene segments. The library is compatible with standard short-read sequencing platforms.

Why does full-length matter? Because many epitope prediction algorithms, including ImmuneWatch DETECT, perform better with complete receptor sequences. Partial CDR3 data limits predictive accuracy; full-length data maximizes it.

Sequencing Multiplexed libraries are sequenced to generate both enriched TCR/BCR transcripts and corresponding whole-transcriptome data, all linked at single cell resolution.

“Our team tested the PythoN dissociator across multiple human tumor tissues and found we were able to extract a significantly higher number of live immune cells compared to a competing tissue dissociation product.”

— Dr. Andrew Shuparski, Scientist II, Immunai, USA

Phase 2: Initial Data Processing (Singleron)

Raw sequencing data is processed through the CeleSCOPE® bioinformatics pipeline, a free command-line tool available on GitHub.

Transcriptome Alignment Paired-end FASTQ files are mapped to the reference genome, generating gene expression matrices for downstream analysis.

Clonotype Assembly Full-length TCR clonotype sequences are reconstructed, with chain pairing information preserved. This is where Singleron’s capture efficiency really shows: the platform achieves mouse TCR α-β chain productive pairing rates exceeding 80% and human BCR IgH-IgK/IgL pairing rates exceeding 85%. High pairing rates mean more complete data for every cell analyzed.

Quality Control Reporting CeleSCOPE generates comprehensive QC reports detailing cellular capture rates, median UMIs per cell, sequence diversity metrics, and chain pairing statistics. This gives you confidence in your data before moving to advanced analysis.

Phase 3: Functional Annotation with ImmuneWatch DETECT

With high-quality clonotype data in hand, the workflow transitions to epitope specificity prediction.

Clonotype Import Assembled TCR sequences are exported from CeleSCOPE and imported into ImmuneWatch DETECT. The tool runs locally and processes a full repertoire in under two minutes on a standard laptop. No GPUs or cloud connectivity required.

Antigen-Specificity Prediction ImmuneWatch DETECT matches input TCR sequences against IMWdb, ImmuneWatch’s curated database of over 2,200 tracked TCR-epitope pairs spanning viral, bacterial, autoimmune, and oncological targets. The algorithm applies probabilistic modeling for epitope specificity prediction providing insights into which epitopes each clonotype is likely to recognize.

Critically, the platform quantifies confidence in each epitope specificity prediction. Not every TCR can be annotated with high certainty, and DETECT is transparent about this, flagging predictions where insufficient evidence exists.

Longitudinal and Differential Tracking For studies involving multiple timepoints or treatment conditions, DETECT enables tracking of antigen-specific clonotypes across samples. You can monitor, for example, how SARS-CoV-2-specific T cells expand or contract following vaccination, or identify tumor-reactive clonotypes that persist after immunotherapy.

Multimodal Integration This is where the integrated workflow delivers its full value. Epitope annotations are merged with full-length gene expression data from the Singleron pipeline, linking TCR specificity directly to transcriptional phenotype. You can identify which antigen-specific clonotypes are exhausted, which are effector-like, and which show stem-like memory characteristics—all at single cell resolution.

“Together with ImmuneWatch we were able to assess vaccine-induced T-cell signatures in our unique longitudinal sample collection of cancer patients. Based on antibody titers, it has previously been shown that cancer patients induce a weaker response towards COVID-19 vaccines. T-cell immunity in this context is far less characterized and understood. Thanks to the ImmuneWatch DETECT platform we were able to take a deep dive into the SARS-CoV-2-specific T-cell response of our cancer patient cohort.”

— Prof. Dr. Eva Lion & Prof. Dr. Timon Vandamme, University of Antwerp – University Hospital Antwerp 

Why Full-Length TCR Data Matters for Epitope Prediction

Most TCR epitope prediction methods focus on the CDR3 region, which is the hypervariable loop that makes primary contact with the peptide-MHC complex. Considering that the CDR3 sequence largely determines antigen specificity, that makes sense.

However, CDR3 isn’t the whole story. The germline-encoded CDR1 and CDR2 loops also contribute to peptide-MHC recognition, particularly for interactions with the MHC molecule itself. V-gene identity can influence binding geometry, and even framework regions can modulate receptor flexibility and docking angle. (1, 2)

ImmuneWatch DETECT leverages this by incorporating V-gene and J-gene information alongside CDR3 sequences. The sCircle® workflow provides exactly this: complete V(D)J sequences, including accurate gene segment assignments, rather than CDR3 fragments alone.

The result is higher annotation confidence and broader epitope coverage. If your upstream data is incomplete, your downstream predictions will be too.

How ImmuneWatch DETECT Delivers Explainable Predictions

Black-box AI algorithms have no place in translational immunology. When you’re trying to understand clinical efficacy or making decision about therapeutic development, you need to understand why the algorithm made a particular prediction. ImmuneWatch DETECT addresses this with explainable, transparent artificial intelligence. For each epitope annotation, the platform provides:

  • Confidence scores based on probabilistic modeling
  • Reference TCRs from IMWdb that informed the prediction

This transparency enables critical evaluation. You can assess whether a prediction is driven by many high-quality reference sequences or a handful of marginal matches. You can trace annotations back to their source data. And you can prioritize candidates for experimental validation based on prediction confidence.

Key Advantages of the Integrated Solution

High Capture Efficiency and Productive Pairing

Singleron’s poly-A-based capture technology delivers robust transcriptome coverage alongside targeted immunoreceptor enrichment. Productive chain pairing rates—exceeding 80% for mouse TCR α-β and 85% for human BCR IgH-IgK/IgL—ensure that most captured cells yield complete, analyzable receptor sequences.

Best-in-Class Epitope Annotation

ImmuneWatch DETECT bypasses the limitations of standard diversity metrics by applying the IMMREP23 benchmark-winning algorithm to predict antigen specificity. (3) With over 2,200 tracked epitopes—and new targets added monthly across cancer, autoimmune, viral, and bacterial domains—the platform accelerates antigen discovery, immune monitoring, and vaccine development. ImmuneWatch DETECT has been validated through external experiments, MLOps best practices, patient-level confirmation, and 14+ peer-reviewed user publications. (4)

Seamless Integration, Actionable Output

The workflow is designed for compatibility. CeleSCOPE outputs feed directly into ImmuneWatch DETECT, and annotated results integrate back with single cell expression data. No format conversions, no data wrangling—just a clear path from raw reads to biological insight.

Resources

  1. (1) Jokinen, E., et al. (2024). “EPIC-TRACE: predicting TCR binding to unseen epitopes using attention and contextualized embeddings.”
  1. (2) Montemurro, A., et al. (2024). “NetTCR 2.2 – Improved TCR specificity predictions by combining pan- and peptide-specific training strategies.” eLife.
  1. (3) Nielsen, M., et al. (2024). “Lessons learned from the IMMREP23 TCR-epitope prediction challenge.” ImmunoInformatics.
  1. (4) Meysman, P., et al. (2026). “How Reliable Are Antigen-Specificity Annotations for T-Cell Receptors? An overview of the state-of-the-art and validation of computational tools for TCR-antigen specificity annotations.” ImmuneWatch White Paper, available at https://www.immunewatch.com/news/white-paper-tcr-antigen-specificity-annotation-reliability-detect

Ready to Move from Sequence to Specificity?

Office-Consultation-Discuss-Talk

Ready to Move from Sequence to Specificity?

Contact us to discuss sample requirements, experimental design, and service options. Reach out for a demo or advice on database coverage, custom training datasets for proprietary epitopes, and integration with your existing analysis pipelines.  Whether you’re working with fresh tissue, cryopreserved cells, or established cell lines, we provide comprehensive protocols and hands-on support.

 

Contact us now

A post by Samantha Langer

Samantha is defined by her journey from the lab bench to strategic leadership. Her scientific foundation, rooted in a PhD in Molecular Biology and hands-on experience as an Application Scientist, provides a critical advantage. She has supported researchers across complex areas like oncology, hematology, and niche fields like gravitational biology, giving her a comprehensive view of scientific challenges.

As Senior Marketing Manager, she directs the team and drives the strategic vision for the European market. Samantha is focused on creating and managing digital campaigns that ensure maximum market recognition and commercial success. Her mission is to effectively bridge Singleron’s innovative technologies with the researchers who rely on them to achieve their next major scientific breakthrough.

Check out our latest blog posts

Learn more
Singleron Publications 2024

2024 in Review: A Year of Remarkable Scientific Breakthroughs 

What a year 2024 has been for the scientific community – in particular in the field of immuno-oncology! Researchers using Singleron’s products have collectively published over 150 studies this year, each one pushing the boundaries of our understanding of diseases and molecular mechanisms. We are immensely proud to celebrate the dedication and success of these brilliant scientists whose work continues to inspire and drive progress in the field. 

Read more