Introduction to Serotypes

What are Serotypes?

A serotype is defined by specific amino acid positions that determine how antibodies recognize and react with HLA molecules. Serotypes provide a systematic way to classify HLA alleles based on their molecular structure, offering more precise matching capabilities than traditional serological methods.

Specific amino-acid residues in the membrane-distal domain form the epitopes that antibodies recognize.
Different combinations of residues at these positions create distinct serological specificities.
This lets alleles that lack traditional serological typing data be classified, and supports more precise prediction of antibody reactions and transplant compatibility.

A serotype is named after a representative prototype allele: usually the lowest-numbered or most common two-field allele with that specificity, written without the colon (HLA-A*02:01 gives A-0201 while proposed, A0201 once WHO-accepted). Many other alleles map to the same serotype, so an allele's own number need not match the serotype it belongs to.

The serological hierarchy

HLA serological specificities are organised from least to most specific. The levels are nested: every serotype falls inside a split, and every split inside a broad antigen.

Broad antigen

least specific

A9, B5, A10

The original, broadly defined specificity: one antibody reactivity that covers a whole family of related molecules.

Typing: Serology or low-resolution PCR (SSO/SSP)

Split antigen

more specific

often just called “the antigen”

A23, A24, B51

A narrower specificity that splits a broad into more defined reactivities.

Typing: Serology or intermediate-resolution PCR (SSO/SSP)

Serotype

most specific

A2301, A2403, B0702

The finest specificity, set by the exact epitope-determining residues that distinguish alleles within a split. Assigning it requires allele-level sequence.

Typing: Allele-level sequencing (NGS, long-read / nanopore)

Naming: A hyphen marks one not yet WHO-approved (A-0201); the hyphen is dropped once approved (A0201).

Associated antigen (often called the serologic equivalent)

The WHO term for the serological specificity assigned to an allele: what it is expected to type as. It is not a separate serologic level; the label can sit at the broad, split, or serotype level, depending on how finely the allele's specificity has been resolved. The 2026 WHO report accepted many HATS serotypes as associated antigens.

Serotypes don't replace broad or split antigens. Each remains a valid serologic equivalent, so you can report whichever level your typing supports.

Bw4 and Bw6 are separate public epitopes

Bw4 and Bw6 are public epitopes shared across most HLA-B antigens (and a few HLA-A). They form their own codominant system and sit outside the broad / split / serotype hierarchy, so an allele can carry a Bw4 or Bw6 assignment in addition to its associated antigen.

Serotypes and eplet matching

Serotypes complement epitope-based (eplet) matching, which estimates compatibility from the load of mismatched eplets between donor and recipient. A serotype collapses an allele's epitope variation into a single, named serological specificity defined by the same membrane-distal, antibody-accessible residues (DEP), giving most alleles a standardized antigen a lab can use without specialized software, allele-level typing, or eplet expertise, and remaining complementary to eplet matching where it is performed.

An evolving classification

Serotype definition is an ongoing effort. The 2024 update validated 20 new antigens against patient antibody data, and the 2025 study added epitope positions 103 and 109, allowing nearly all common HLA-B antigens to be defined at the most precise (Full) level, and proposed 18 further antigens across HLA-B and HLA-DP. Newly accepted specificities are folded into WHO nomenclature (most recently the 2026 report) and the IPD-IMGT/HLA database, so the set of serotypes continues to grow with each release.

Example: the A9 family

Broad Antigen

Split Antigen

A23

Split Antigen

A24

Serotype

A2301

Serotype

A2304

Serotype

A2402

Serotype

A2403

Serotype

A2404

Serotype

A2405

Serotype

A2408

Serotype

A2410

Serotype

A2414

Serotype

A2423

Serotype

A2424

Serotype Assignment Process

The reference patterns were established by correlating the serological reactivity of allo-antisera and monoclonal antibodies with specific amino acid residues. New HLA alleles are then classified by comparing their sequences against these patterns at the defining (F and S) positions, and the sequence-based predictions are confirmed, and new specificities discovered, using single antigen bead (SAB) assays.

Sequence Analysis
Each allele's amino acid sequence is examined at designated critical positions based on F (Full) and S (Serotype) tier classification.
Pattern Comparison
Sequences are compared against prototype patterns to determine their closest serotype match.
Assignment Scores
Each allele receives a qualifier score reflecting how well it matches the reference pattern (see Qualifier Scores below).
Serological Confirmation & Discovery
Predicted antigens are validated against single antigen bead (SAB) solid-phase assays, where each bead carries one prototype allele's antigen. Mean fluorescence intensity (MFI) values from large patient-serum datasets (over 13,000 sera in the 2024 update) are compared between bead pairs: high correlation indicates the same antigen, while low correlation flags a distinct antigen, letting the method confirm computationally predicted serotypes and identify new specificities not previously listed. Ambiguous outliers are resolved on live cells by flow cytometric crossmatch (FXM): this also separates genuine new specificities from bead artifacts, since some beads give high readings from denatured or cryptic protein that is not present on intact cells. Serotypes still awaiting this confirmation remain valid sequence-based predictions, marked with the InSilico score.

How one amino acid can change a serotype

Serotypes are defined by a few amino-acid positions where antibodies bind, the defining positions shown on each serotype page. A change at one of these can make two near-identical molecules into different antigens, while other changes have no effect. The 2025 study pinpointed positions 103, 109 and 167 in HLA class I.

A change that matters (position 103)

B*35:01

103 L

→

B35

B*35:12

103 V

→

B-3512

Different residue, different antigen.

A change that doesn't (position 167)

B*15:12

167 G

→

B76

B*15:14

167 S

→

B76

Different residue, same antigen.

Reading the antibody-correlation plots

Each point is one patient's serum, plotting how strongly it reacts (mean fluorescence intensity, MFI) with two single antigen beads. Sera positive on both beads (black) line up along the regression line when the beads carry the same antigen (high R²); sera positive on only one bead cluster along that axis (blue or red). When the beads are different antigens, the points scatter off the line (low R²). The dashed lines mark the MFI 1000 positivity cutoff. The examples below are illustrative.

Same serotype (R² ≈ 0.95)

B*07:02 and B*07:04 are both serotype B0702

Distinct serotypes (R² ≈ 0.20)

Scatter off the line: split by residue 163 (B4002 vs B4005)

Schematic of the method, following the MFI scatter convention used in the HATS validation (representative points, not individual patient data). Osoegawa et al. 2025

Qualifier Scores

F (Full)

The allele matches the serotype's reference pattern at all of its defining amino-acid positions — the most complete match.

Example: Complete alignment with prototype allele pattern using all polymorphic residues.

S (Serotype)

The allele matches at the key positions that define the serotype's serological specificity, though not at every position of the full pattern.

Example: Alignment at key epitope-determining positions for serological recognition.

I (Incomplete)

The allele differs from the closest serotype's pattern by a single amino acid, so the match is partial rather than exact.

U (Unassigned)

The allele differs at two or more defining positions and was not matched to a serotype.

InSilico

The serotype is predicted computationally from the sequence and has not yet been validated by serological testing (single-antigen-bead or serum reactivity).

Blank

A blank score marks a null or non-expressed allele (e.g. names ending in N). It produces no cell-surface protein, so it has no serotype.

The score describes how completely an allele's sequence matches the serotype's defining residues (F = all, S = the key subset, I = one mismatch, U = no match) and whether the assignment has been serologically validated (InSilico = predicted from sequence only).

Note on older publications: in the original 2022 method, “S” meant “Short” and “SC” meant “Short-Crossreactive” (both forms of an incomplete match). The current classification re-uses S to mean Serotype, a confident assignment, so if you are reading the 2022 paper, don't confuse the two.