Draft status: This content reflects the current working spec and may be revised.

1. What is PLSC?

Phenotype List String Code (PLSC) pairs a PL-String with its namespace and version (or date) so that every expression is self-contained and traceable.

<namespace>#<version or date>#<pl-string>
  • Namespace — who defines the vocabulary (e.g., who, optn, et, nmdp)
  • Version or date — a release tag (preferred) or ISO date (YYYY-MM-DD)
  • PL-String — a phenotype expression using the operators below

Use a release version when available (e.g., 3.61.0). If not, use the date the reagent/result/interpretation was produced.

2. Namespaces

A namespace represents a governed, versioned code system (or a set of systems with one base). Examples:

  • who — Base: IPD-IMGT/HLA release; Extensions: WHO antigen names, NMDP MAC codes for ambiguity, and other well-documented, versioned tables.
  • optn — OPTN antigen/epitope codes used in U.S. allocation workflows.
  • et — Eurotransplant antigen/match determinant tables.
  • nmdp — registry codes (including MACs), aligned to specific IMGT/HLA releases.

Gene family namespace

Each gene family namespace represents one or more code systems. When more than one code system is used, a base system is designated which is then extended by the other code systems.

Code Systems Comprising who Namespace - who — Base: IPD-IMGT/HLA release; Extensions: WHO antigen names, NMDP MAC codes for ambiguity, and other well-documented, versioned tables. - optn — OPTN antigen/epitope codes used in U.S. allocation workflows. - et — Eurotransplant antigen/match determinant tables. - nmdp — registry codes (including MACs), aligned to specific IMGT/HLA releases.

Rules

  • A PLSC must specify exactly one namespace and one version (or date).
  • No mixing of namespaces within a single PL-String.
  • Each implementation should publish its governance, release process, and documentation for users.

3. PL-String grammar

PL-String encodes phenotypes (not genotypes). It represents the presence or possible presence of proteins/antigens and heterodimer pairing for class II. It does not carry chromosomal phase, zygosity, or locus order.

3.1 Delimiters and precedence

Precedence Delimiter Name Meaning (phenotype context) Typical context
1 + AND All listed molecules/antigens are present (e.g., on a bead; in a result) All contexts
2 ~ Heterodimer Exactly two elements form one heterodimer (class II α~β) DQ, DP, DR (protein/antigen)
3 % Inclusive OR One or more of the listed items may be present Antigen lists; multi-reactivity
4 / Exclusive OR Exactly one of the listed items is present (ambiguity) Protein-level ambiguity (e.g., MAC)

Notes:

  • + binds most tightly (1), / loosest (4) for unambiguous parsing.
  • % (inclusive OR) is primarily for antigen ambiguity/association.
  • / (exclusive OR) encodes protein ambiguity (e.g., unresolved allele).
  • ~ is only valid for appropriate α~β pairs per the chosen namespace.

3.2 Basic rules

  • No loci/phase semantics: PL-Strings do not assert chromosomal phase, cis/trans, zygosity, or locus order; DPA1*01:04~DPB1*02:02 is equivalent to DPB1*02:02~DPA1*01:04.
  • Slash locus constraint: Terms on either side of / must be the same locus/category within the namespace (e.g., HLA-A*01:01/HLA-A*01:02 is valid; mixing A with B is not).
  • Heterodimers: ~ may only connect permitted α~β pairs as defined by the namespace (e.g., DQA1~DQB1, DPA1~DPB1, DRA~DRB3/4/5 or their antigen equivalents).
  • No mixed namespaces inside a single PL-String.
  • Whitespace is not significant; implementers should normalize.

3.3 Examples

Reagent (protein ambiguity on a bead)
HLA-A*01:01/HLA-A*01:02+HLA-A*02:01

A*02:01 is present; A*01:01 or A*01:02 (but not both) is present.

Result (antigen-level inclusive ambiguity)
A0101%A0102+A2402

A2402 present; A0101 and/or A0102 may be present.

Class II heterodimers (protein level)
HLA-DQA1*05:01~HLA-DQB1*02:01%HLA-DPA1*01:03~HLA-DPB1*04:02

One or both heterodimers may be present.

Constrained heterodimer variants
HLA-DPA1*01:04~HLA-DPB1*02:01%HLA-DPA1*01:04~HLA-DPB1*04:02

One or two heterodimers; DPA1*01:04 is always present.

Multiple loci (phenotype list)
HLA-A*01:02+HLA-A*02:01+HLA-B*07:02

A*01:02, A*02:01, and B*07:02 present.

Implementations should reject invalid pairs like DQA1*05:01~DPB1*04:02 based on namespace pairing rules.

4. PLSC construction

Syntax

namespace#version_or_date#plstring

Version: Prefer authoritative release tags (e.g., 3.61.0).
Date: If no release applies, use ISO YYYY-MM-DD. The date reflects when the reagent/result/interpretation was produced—not sample collection/export.

4.1 Examples

  • IPD-IMGT/HLA release bound protein list:
    who#3.61.0#HLA-A*01:01/HLA-A*01:02+HLA-A*24:02
  • Class II heterodimers with antigen mapping:
    who#2025-11-02#DQA1*05:01~DQB1*02:01%DPA1*01:03~DPB1*04:02
  • Ambiguous MAC usage (namespace-governed):
    nmdp#2025-11-02#HLA-DQB1*03:01/297+HLA-DQB1*06:02

5. Validation checklist

  1. Namespace present and recognized.
  2. Version or date present and well-formed.
  3. No mixed namespaces within <pl-string>.
  4. Operators parse with precedence: + > ~ > % > /.
  5. Slash locus constraint holds on every /.
  6. Heterodimer pairs are permitted by the namespace.
  7. Atoms exist in the declared namespace release/date.

6. Embedding in exchange formats

  • FHIR: Use as a Coding with system (e.g., http://plstring.org), version (namespace release or date), and code (the full PLSC string).
  • HAML: Store as the code value for reagent/result/interpretation elements.
  • No mixing of namespaces inside one code value; use separate codings for different namespaces.

FHIR example

<valueCodeableConcept>
  <coding>
    <system value="http://plstring.org"/>
    <version value="1.0"/>
    <code value="who#3.61.0#HLA-DQA1*05:01~HLA-DQB1*02:01+HLA-DPB1*04:01"/>
  </coding>
</valueCodeableConcept>

7. Comparison to GL-String (at a glance)

Concept GL-String (genotype) PL-String (phenotype)
Focus Alleles/genotypes & phase Proteins/antigens & heterodimers
Locus delimiter ^, |, ? (various) Not used
Exclusive OR / (allele ambiguity) / at protein level
Inclusive OR (N/A) % at antigen level
Heterodimer ~ for phase ~ for α~β pairing (class II)
Composition + + (reagent contents; multi-reactivity)
Mix namespaces Allowed per implementer Disallowed within a single PL-String

8. Frequently asked

Can I encode epitopes/eplets?
Not in this version. Once stable, widely adopted nomenclatures are available, the grammar can extend to epitope atoms (e.g., OPTN DPB1 unacceptable epitopes, eplet registries).

What about non-classical loci (E, F, G) or MICA/MICB/ABO?
Permitted if your namespace publishes governed vocabularies; apply the same operators and rules.

This page supersedes earlier drafts and aligns operator semantics (+, ~, %, /), constraints (no mixed namespaces; heterodimer pairing rules), and PLSC binding (namespace + version/date) with the current manuscript. It replaces the older GL-centric placeholder text.