HLA Typing and Immune Repertoire Sequencing Services: From High-Resolution HLA Genotyping to TCR/BCR Profiling

Meta Intent: A practical research guide for choosing HLA typing, TCR/BCR immune repertoire sequencing, or a combined immunogenomics strategy based on the biological question, required resolution, sample type, and evidence needed.

The Immune System Has More Than One Molecular Barcode

Immune-genetics projects often begin with a deceptively simple request: "profile the immune system." Yet that request can refer to three different molecular layers. HLA typing identifies inherited variants that shape antigen presentation. TCR and BCR sequencing measures rearranged adaptive-receptor populations in a sample. KIR typing adds a partly independent layer of natural killer cell genetics and HLA-ligand context.

These layers are connected, but they are not interchangeable. HLA genotyping cannot show which lymphocyte clones have expanded. Repertoire sequencing cannot reconstruct a complete HLA genotype. A diversity score cannot establish antigen specificity, and an HLA–KIR genotype does not directly measure NK-cell function. The right study therefore begins by defining the biological object to be observed rather than selecting every available assay.

This guide organizes HLA Typing, TCR & BCR Sequencing, and related sequencing options into a decision framework. It explains what each layer can support, where interpretation commonly fails, and when a combined design creates information that neither assay can provide alone.

Figure 1: Three layers of immune-genetic evidence. HLA defines inherited antigen-presentation context, TCR/BCR repertoires describe adaptive clonal composition, and HLA–KIR combinations add an NK-cell regulatory layer. The layers interact biologically but answer different analytical questions.

Start With the Evidence Claim, Not the Assay Name

Before choosing a workflow, write the intended conclusion as one sentence. A useful statement specifies the comparison, the biological unit, and the evidence expected. Examples include:

• "Determine whether HLA allele frequencies differ between two research cohorts."
• "Measure whether selected TCR clonotypes expand between baseline and a later time point."
• "Compare BCR clonal diversity between paired tissue compartments."
• "Annotate donor and recipient HLA alleles for a transplantation research dataset."
• "Explore whether HLA ligand groups and KIR gene content stratify an NK-cell phenotype."

Each claim points toward a different primary measurement. Cohort-level inherited variation points to HLA genotyping. Longitudinal clonal change points to repertoire sequencing. Receptor pairing or receptor-to-cell-state linkage may require single-cell analysis. A hypothesis about NK-cell regulation may require both HLA and KIR information, ideally paired with an independent functional or phenotypic assay.

This step also exposes questions that sequencing alone cannot answer. Observing an expanded TCR clonotype does not prove what antigen it recognizes. Finding an HLA allele associated with a phenotype does not prove mechanism. Detecting a KIR–ligand combination does not establish receptor expression or NK-cell activity. Those are downstream validation questions, not reasons to overinterpret the sequencing layer.

A Decision Map for HLA, TCR/BCR, and Combined Designs

The most efficient selection logic is to ask what varies in the study.

The table is a starting point, not a universal protocol. Study design still determines whether the unit of inference is a person, specimen, tissue, time point, sorted population, or cell. If biological replication and metadata are weak, adding another sequencing assay will expand the data volume without strengthening the claim.

Define the smallest sufficient assay package

The word "comprehensive" should describe evidence coverage, not the number of assays purchased. A smallest sufficient package contains every measurement required for the primary claim and excludes measurements that have no predefined analytical role. For an HLA association study, this may mean a consistent locus panel, participant-level metadata, and a prespecified statistical model. For longitudinal immune monitoring, it may mean one receptor chain, carefully paired time points, and a reproducible clonotype pipeline. For an antigen-presentation hypothesis, it may require HLA typing plus repertoire sequencing and a separate specificity assay.

Three planning questions expose unnecessary scope quickly. First, which output enters the primary analysis? Second, what decision would change if an optional assay produced a positive or negative result? Third, what validation is required before the result can support the stated conclusion? If an output has no role in any of these answers, it is likely exploratory. Exploratory measurements can still be valuable, but they should be labeled, powered, and interpreted separately from the primary endpoint.

This discipline also improves sample allocation. Limited DNA, RNA, or viable cells should be reserved according to evidentiary priority. Consuming the best material in a broad exploratory library can compromise the assay that actually tests the hypothesis. A staged plan—feasibility check, primary assay, then conditional extension—often produces a stronger dataset than running all methods in parallel from the beginning.

Figure 2: Assay-selection decision logic. A question-led decision tree separates inherited HLA variation, bulk clonal architecture, paired receptor analysis, and HLA–KIR hypotheses before platform selection.

HLA Typing: Resolution Is a Reporting Decision and a Sequencing Decision

HLA loci are exceptionally polymorphic. Closely related alleles may differ within peptide-binding exons, elsewhere in coding sequence, in introns, or in untranslated regions. This creates two connected choices: which regions must be observed and how specifically results must be reported.

Older shorthand such as "two-digit" and "four-digit" persists, but field-based nomenclature is clearer. In an allele such as HLA-A*02:01:01:01, the first field identifies an allele group, the second distinguishes a specific HLA protein sequence, the third captures synonymous coding differences, and the fourth distinguishes noncoding sequence differences. Expression suffixes can convey additional information. High resolution is commonly associated with at least second-field molecular reporting, but "full-gene sequencing" describes sequence coverage rather than automatically guaranteeing an unambiguous four-field call.

This distinction matters. A project that only needs allele groups for broad population stratification may not benefit from sequencing every intron. A disease-association study may need consistent second-field calls across a large cohort. A project concerned with phasing, novel variants, noncoding variation, or extended haplotypes may justify full-gene or long-read coverage. The analytical specification should therefore state:

• loci to be typed;
• regions covered at each locus;
• requested reporting field or acceptable allele groups;
• how G groups, P groups, null alleles, and unresolved ambiguities will be handled;
• reference database release and analysis software version;
• minimum evidence for a novel or discordant call.

For a detailed comparison of molecular methods, see molecular HLA typing technologies. Projects that require locus-focused sequencing can also be framed through Targeted Region Sequencing or Amplicon Sequencing Services, depending on target length, locus complexity, and the need for phase information.

The dedicated matrix article High-Resolution HLA Typing by NGS will examine class I and class II loci, assay architecture, allele nomenclature, ambiguity handling, and result interpretation in depth.

Figure 3: HLA resolution from locus to phased full-gene sequence. The illustration distinguishes first-field allele groups, second-field protein-level specificity, synonymous coding variation, noncoding variation, and physical phasing across an HLA gene.

Short Reads, Long Amplicons, and Long Reads Solve Different HLA Problems

Platform choice should follow the unresolved ambiguity in the research question. Short-read targeted NGS can support high-throughput typing when validated amplification, balanced locus coverage, and appropriate algorithms are available. However, short fragments may not directly phase distant polymorphisms. Statistical inference or allele databases may resolve many combinations, but difficult or novel haplotypes can remain ambiguous.

Longer molecules can bridge distant variants and support direct allele phasing. A Long Amplicon Analysis design may be appropriate when the target can be amplified intact and the study needs consensus sequences across a defined locus. Nanopore Target Sequencing can support long-range targeted analysis where molecule length and haplotyping are central. These benefits do not eliminate the need to monitor allele dropout, amplification imbalance, homopolymers, mapping ambiguity, and reference incompleteness.

Orthogonal confirmation should be driven by risk, not habit. Sanger Sequencing can resolve a defined discrepant position or confirm a novel variant in a focused region, but it is not a substitute for phase-aware full-gene evidence. A defensible confirmation plan predefines which events trigger review: low coverage, allele imbalance, novel combinations, discordance with prior typing, unexpected homozygosity, or software disagreement. The general principles in Sanger validation of NGS results help distinguish targeted confirmation from redundant resequencing.

HLA in Transplantation and Cord Blood Research

HLA matching is central to transplantation research, but the loci, matching thresholds, and interpretation rules depend on the graft source and study context. A single phrase such as "10/10 matching" should never be copied between unrelated donor, cord blood, solid-organ, or cellular research projects without defining what was counted and at what resolution.

Cord blood illustrates why context matters. Cord blood units have historically been selected under different matching tolerances than adult unrelated donors, but unit quality and cell dose remain important alongside HLA compatibility. High-resolution typing can strengthen inventory characterization and confirmatory research, yet it does not replace the other variables required in a unit-selection framework. Similarly, the presence of a permissible mismatch in one research model should not be generalized to another population or protocol.

A project brief for cord blood should separate at least four questions:

1. Which loci and resolution are needed at banking or inventory characterization?
2. Which specimen will be retained for confirmatory testing and identity checks?
3. How will sample provenance, chain of custody, and metadata be recorded?
4. Which matching algorithm and research endpoint will be used downstream?

The matrix article HLA Typing for Cord Blood Banking and Transplantation will address these design issues, including the distinction between unit characterization and downstream matching decisions. CD Genomics provides sequencing support for research use; treatment decisions and regulated histocompatibility testing remain within the appropriate clinical and accredited laboratory framework.

Figure 4: Cord blood evidence chain. The diagram separates unit identity and sample traceability, HLA typing, matching analysis, and non-HLA unit attributes so that sequencing evidence is not treated as the entire selection decision.

Immune Repertoire Sequencing Measures Rearranged Populations, Not Germline HLA

TCR and BCR receptor diversity is generated by V(D)J recombination and junctional modification. B-cell repertoires are further shaped by somatic hypermutation and class-switch recombination; class switching changes the antibody constant region and effector isotype rather than the antigen-binding V(D)J sequence. Repertoire sequencing captures rearranged receptor molecules from the sampled cells or nucleic acids. The resulting data can describe clonal composition, V/J gene usage, CDR3 sequence properties, diversity, evenness, clonality, overlap, and longitudinal expansion.

The output is shaped by what enters the tube. Whole blood, peripheral blood mononuclear cells, sorted lymphocytes, tumor tissue, and adjacent tissue represent different cellular mixtures. DNA-based and RNA-based assays also answer different quantitative questions. Genomic DNA can provide a closer relationship between rearrangements and cell counts for appropriate loci, whereas RNA reflects receptor transcript abundance and can offer high sensitivity but introduces expression-level variation. Neither choice is universally superior.

For RNA-based profiling, extraction quality and transcript abundance affect detectable clonotypes. Projects with limited material may benefit from an Ultra Low RNA Sequencing discussion during feasibility planning, although a receptor-specific library remains distinct from general transcriptome sequencing. Conventional RNA-Seq can provide broader expression context but should not be assumed to deliver the sensitivity, chain coverage, or clonotype quantification of a dedicated repertoire assay.

The focused article TCR and BCR Immune Repertoire Sequencing will cover receptor chains, library strategies, clonotype definition, diversity metrics, and application-specific analysis. Additional background is available in the guides to bulk immune repertoire sequencing and BCR versus TCR sequencing.

Bulk or Single Cell: Decide Whether Pairing and Phenotype Are Necessary

Bulk repertoire sequencing is efficient for measuring population-level clonotype frequencies across many samples. It can support longitudinal tracking, tissue comparisons, V/J usage analysis, and diversity profiling. Its main limitation is loss of native chain pairing and cell-level context. A TRA sequence and a TRB sequence found in the same bulk sample cannot automatically be assigned to the same T cell. The same limitation applies to immunoglobulin heavy and light chains.

Single-cell repertoire sequencing can preserve paired chains and link them to an individual cell. When combined with expression or surface-protein measurements, it can connect receptor identity to a cellular state. That added information is valuable only if the research question needs it. If the goal is to compare dominant TRB clones across 200 longitudinal samples, bulk profiling may provide stronger cohort coverage. If the goal is to recover paired receptors from a rare responding population, single-cell analysis may be essential.

The decision should consider:

• Is native alpha–beta or heavy–light pairing required?
• Must receptor sequences be assigned to cell subsets or transcriptional states?
• Is the sample viable and suitable for single-cell capture?
• Is broad cohort coverage more important than deep cellular context?
• Will the analysis compare cells, specimens, or participants as the inference unit?

The single-cell TCR sequencing overview provides additional context for pairing-aware designs. The correct comparison is therefore not "bulk is basic, single cell is advanced." It is "which data structure is necessary to test the claim?"

Figure 5: Bulk and single-cell repertoire outputs. One sample is split into population-level clonotype frequencies on the left and cell-linked paired receptors with phenotype metadata on the right, emphasizing the different inference units.

Repertoire Metrics Need a Predefined Interpretation Contract

Repertoire data are compositional and sampling-dependent. A larger clone occupies a greater fraction of reads and changes the apparent abundance of every other clone. Richness rises with sampling effort. Diversity metrics collapse different biological structures into a single number. Two samples can have similar Shannon diversity while sharing few clonotypes, and two samples can share a dominant clone while differing substantially in their low-frequency tail.

A useful analysis plan therefore specifies the estimand before choosing a metric. Possible estimands include dominant-clone expansion, total observed richness at standardized sampling effort, within-sample evenness, between-sample overlap, V/J usage shifts, CDR3 property distributions, or B-cell lineage evolution. Each one requires different preprocessing and statistical treatment.

Clonotype definition must also be explicit. Depending on the question, sequences may be grouped by nucleotide identity, amino-acid CDR3, V/J assignment, paired chains, or lineage-aware similarity. Error correction, UMI handling, productivity filters, minimum abundance thresholds, and treatment of ambiguous V/D/J calls can materially change the result. Pipeline and reference versions must therefore accompany every reported metric.

The AIRR Community's MiAIRR framework reinforces this point by structuring metadata from study design and sample collection through nucleic-acid processing, sequencing, data processing, and annotated rearrangements. A reproducible project should preserve these links rather than delivering only a spreadsheet of clonotypes.

HLA and Repertoire Data Become More Informative When the Hypothesis Connects Them

HLA genotype constrains which peptide–HLA complexes may be presented, while antigen processing, expression, and cellular context also affect presentation; TCR sequencing observes receptor sequences in the sampled T-cell population. Combining the two can support HLA-stratified repertoire comparisons, candidate response monitoring, cohort analysis, and prioritization for downstream antigen-specific experiments. It still does not prove that a given receptor recognizes a particular peptide–HLA complex.

A combined study should declare the bridge between layers. For example:

• compare repertoire features among participants carrying a predefined HLA allele;
• test whether a clonotype expansion occurs only in an HLA-compatible subset;
• incorporate HLA genotype as a covariate in a longitudinal repertoire model;
• select candidate receptors for independent peptide–HLA binding or functional assays;
• integrate receptor clonotypes with transcriptomic states through a Multi-Omics Service when multiple molecular layers are genuinely required.

The strongest designs avoid post hoc storytelling. HLA strata, receptor chain, time points, covariates, clonotype definition, and confirmatory criteria should be fixed before inspecting the outcome. Otherwise, the number of possible allele–clonotype combinations creates a large multiple-testing burden and an attractive route to unstable associations.

Figure 6: HLA-to-repertoire hypothesis bridge. The visual links a predefined HLA stratum to receptor sampling, statistical comparison, and independent functional validation, with a visible boundary between association and specificity evidence.

HLA–KIR Typing Adds an NK-Cell Genetics Layer

KIR genes encode activating and inhibitory receptors expressed in variable combinations on NK cells and some T-cell subsets. Their ligands include motifs carried by selected HLA class I molecules. Because KIR and HLA loci reside on different chromosomes and are inherited independently, a person may carry a KIR gene without carrying its corresponding HLA ligand, or vice versa.

This makes combined HLA–KIR analysis useful for hypotheses about NK-cell education, missing-self recognition, infection research, reproductive immunology, disease association, and transplantation biology. It also makes interpretation easy to overstate. Gene presence does not demonstrate expression on a particular NK-cell subset. Ligand assignment is not a direct functional assay. Published transplantation associations can vary with graft source, conditioning, population, endpoint, and the model used to define mismatch.

A research report should distinguish KIR gene content, copy number, allele calls, inferred haplotype, HLA ligand group, and observed phenotype. These variables should not be collapsed into a single "compatible/incompatible" label unless the algorithm and biological model are stated.

The matrix article HLA-KIR Typing will examine KIR locus complexity, HLA-C1/C2 and Bw4 ligand concepts, genotyping approaches, and model-specific interpretation. In the Hub, the practical message is simpler: add KIR typing only when the study contains an NK-cell hypothesis that HLA typing alone cannot test.

Figure 7: HLA–KIR genotype-to-function boundary. KIR gene content and HLA ligand groups feed into a genetic interaction model, while receptor expression and NK-cell activity remain separate measurements requiring independent evidence.

Sample Architecture Determines What Can Be Compared

The same participant can yield different immune profiles from whole blood, PBMCs, sorted cells, marrow, lymphoid tissue, tumor tissue, or adjacent tissue. Collection timing, anticoagulant, processing delay, cryopreservation, extraction method, and cell enrichment can alter the observed repertoire. HLA genotype should be stable across suitable germline DNA sources, but typing quality still depends on DNA integrity, concentration, contamination, and locus-specific amplification.

For HLA typing, record specimen source, DNA extraction method, quality metrics, locus panel, failed loci, repeat criteria, and sample identity controls. For repertoire sequencing, add cell subset, cell count, viability where relevant, nucleic-acid type, input amount, collection time, tissue compartment, treatment or exposure time point, and whether replicates are biological or technical.

Batch design is part of the biology. If all cases are processed in one run and all controls in another, a batch effect is inseparable from group. Longitudinal specimens from one participant should be distributed deliberately, and paired samples should remain identifiable through the pipeline. Controls can include replicate amplifications, defined clonotypes, synthetic spike-ins, complex biological controls, negative controls, and cross-run references, selected according to the failure mode that matters.

A Minimum Defensible Workflow for Combined Immune Profiling

A combined project does not need every possible assay. It needs a chain of evidence with explicit checkpoints.

1. Define the claim. State the comparison, inference unit, primary endpoint, and excluded interpretations.
2. Select the immune layer. Choose inherited HLA, adaptive repertoire, HLA–KIR, or a justified combination.
3. Freeze the sample map. Link participant, specimen, tissue, time point, extraction, and batch identifiers.
4. Specify assay resolution. Define HLA loci and reporting fields; define receptor chains, DNA/RNA input, bulk/single-cell mode, and clonotype rules.
5. Predefine QC gates. Include sample sufficiency, locus coverage, allele balance, repertoire read support, contamination checks, and repeat triggers.
6. Version the references. Record HLA and receptor germline database releases, software versions, and parameter settings.
7. Analyze at the correct unit. Cells and reads are not independent participant replicates. Statistical inference should reflect the study design.
8. Separate association from mechanism. Use orthogonal assays when the claim involves antigen specificity, receptor function, expression, or cell phenotype.

This workflow creates a useful stopping rule. If the claim can be supported by HLA typing alone, do not add repertoire sequencing for decoration. If receptor pairing is unnecessary, do not default to single cell. If a genetic HLA–KIR association cannot be connected to an NK-cell endpoint, state that limitation before data generation.

Figure 8: Minimum defensible immune-profiling workflow. Eight QC-linked stages move from claim definition to assay selection, metadata, sequencing, versioned analysis, inference, and orthogonal validation.

What a Research-Ready Deliverable Should Contain

A useful HLA report contains more than final allele names. It should include loci attempted and completed, coverage or confidence information, ambiguity notation, homozygous versus potentially dropped-allele review, reference database version, software version, and flags for novel or discordant observations. The HLA typing result interpretation guide explains how nomenclature and ambiguity affect downstream use.

A repertoire deliverable should connect raw reads to filtered clonotypes. At minimum, retain raw-data QC, productive and nonproductive rearrangement counts, chain-level summaries, clonotype definitions, frequency tables, V/D/J assignments where applicable, CDR3 sequences, diversity metrics with calculation details, and sample metadata. Figures should be reproducible from machine-readable tables rather than existing only as static images.

For combined analysis, deliver a participant-level key that joins HLA, KIR, repertoire, phenotype, and covariate data without mixing identifiers or exposing unnecessary personal information. Include the statistical model, excluded samples, missingness, batch variables, and multiplicity strategy. A short limitations section should explicitly state what the data cannot establish.

How CD Genomics Supports Question-Led Immune Profiling

CD Genomics supports research projects spanning HLA typing, KIR characterization, and TCR/BCR repertoire sequencing. Project design can be adapted around locus coverage, reporting resolution, bulk or pairing-aware receptor analysis, sample constraints, and downstream data requirements. Related targeted and long-read options can be considered when phase, long-range coverage, or orthogonal confirmation is necessary.

The most productive consultation starts with a concise project brief rather than a platform request. Include the research hypothesis, organism, cohort and group structure, sample type, number of specimens, longitudinal or paired relationships, required HLA loci, receptor chains, desired reporting resolution, primary analysis endpoint, and planned validation. This information allows the sequencing plan to be matched to the evidence claim and reduces avoidable method stacking.

Project Planning Checklist

• What exact conclusion should the project support?
• Is the target inherited genotype, rearranged receptor population, paired receptor, or NK-cell genetic context?
• What is the biological replicate and inference unit?
• Which HLA loci and reporting fields are necessary?
• Which TCR or BCR chains must be captured?
• Is bulk frequency sufficient, or is native chain pairing required?
• Which specimen, tissue, cell subset, and time point will be analyzed?
• Which controls detect amplification bias, contamination, dropout, and batch effects?
• Which database and software versions will be recorded?
• What evidence will trigger repeat testing or orthogonal confirmation?
• Which claims require an independent functional assay?
• How will the Hub and Spoke outputs be joined without duplicating analyses?

Frequently Asked Questions

CD Genomics services are provided for research use only and are not intended for diagnostic procedures or individual treatment decisions.

References:

1. Mayor NP, Robinson J, McWhinnie AJM, et al. HLA Typing for the Next Generation. PLOS ONE. 2015;10(5):e0127153. DOI: 10.1371/journal.pone.0127153.
2. Douillard V, Castelli EC, Mack SJ, et al. Approaching Genetics Through the MHC Lens: Tools and Methods for HLA Research. Frontiers in Genetics. 2021;12:774916. DOI: 10.3389/fgene.2021.774916.
3. Vander Heiden JA, Marquez S, Marthandan N, et al. AIRR Community Standardized Representations for Annotated Immune Repertoires. Frontiers in Immunology. 2018;9:2206. DOI: 10.3389/fimmu.2018.02206.
4. Trück J, Eugster A, Barennes P, et al. Biological Controls for Standardization and Interpretation of Adaptive Immune Receptor Repertoire Profiling. eLife. 2021;10:e66274. DOI: 10.7554/eLife.66274.
5. Gupta AO, Wagner JE. Umbilical Cord Blood Transplants: Current Status and Evolving Therapies. Frontiers in Pediatrics. 2020;8:570282. DOI: 10.3389/fped.2020.570282.
6. Martínez-Losada C, Martín C, Gonzalez R, et al. Patients Lacking a KIR-Ligand of HLA Group C1 or C2 Have a Better Outcome after Umbilical Cord Blood Transplantation. Frontiers in Immunology. 2017;8:810. DOI: 10.3389/fimmu.2017.00810.