Inborn Errors of Immunity (IEI)

name: Inborn Errors of Immunity
display_name: Inborn Errors of Immunity (IEI)
creation_date: "2026-06-13T00:00:00Z"
description: >-
  Inborn errors of immunity (IEI) are Mendelian disorders caused by germline
  variants in genes required for immune development, regulation, or effector
  function. The IUIS Expert Committee's 2024 classification — the current
  edition, superseding the 2022 update — describes 559 IEI across 508 genes and
  17 phenocopies, of which 67 monogenic defects and 2 phenocopies are new since
  2022, organized into ten phenotypic tables spanning recurrent infection,
  immune dysregulation, autoinflammation, complement deficiency and bone marrow
  failure. Each condition is individually rare, but together they are not: a
  recent US study put the incidence of IEI at 6 per 10,000 people.
grouping_basis:
- SHARED_MECHANISM
- CLINICAL_CONVENTION
grouping_rationale: >-
  Grouped on the defining monogenic-immunity mechanism: every member is a
  germline inborn error disrupting the immune system, as recognized by the
  International Union of Immunological Societies (IUIS) phenotypic
  classification. Members are kept as separate Disease entries (and further
  organized into IUIS-table sub-groupings) because causal gene, inheritance,
  dominant clinical phenotype (infection vs dysregulation vs autoinflammation),
  and organ involvement differ. This grouping points down to nested IUIS-table
  groupings rather than listing all 500-plus entities flatly. The sub-groupings
  follow the IUIS table structure one-to-one, so an entity's table assignment —
  recorded per entry as `classifications.iuis_category` — determines exactly one
  sub-grouping, and the sub-groupings partition the members rather than
  overlapping. Note this holds for disease entities, not for genes: IUIS
  explicitly enters one gene in two tables when different variant classes give
  different phenotypes (RELA, OTULIN, NLRP1, ZAP70, IRAK4, IL6ST, STAT5B,
  FOXN1), so a gene-keyed query across this tree can legitimately return two
  sub-groupings.
  Somatic IEI phenocopies (IUIS Table 10), secondary immunodeficiencies (e.g.,
  HIV/AIDS), and incidental immunodeficiency features in non-IEI syndromes are
  out of scope.
mappings:
  mondo_mappings:
  - term:
      id: MONDO:0003778
      label: inborn error of immunity
    mapping_predicate: skos:exactMatch
    mapping_source: MONDO
    mapping_justification: >-
      The grouping concept is the MONDO inborn error of immunity class
      (MONDO:0003778). Nested IUIS-table sub-groupings are clinical unions
      below this umbrella; several lack a valid MONDO grouping term (see
      research/grouping_mondo_gaps.md and sub-grouping notes).
    consistency:
    - reference: MONDO
      consistent: CONSISTENT
      notes: >-
        Umbrella maps exactly to MONDO:0003778. IUIS sub-table groupings
        are dismech-curated unions and are not required to is-a-subsume
        under this term in MONDO.
membership_criteria:
- description: >-
    A disorder is an inborn error of immunity if it is a Mendelian germline
    disorder with a causal variant in a gene required for immune development,
    regulation, or effector function, as assigned to an IUIS IEI phenotypic
    table. Excludes acquired immunodeficiency and somatic IEI phenocopies.
  criteria_semantics: NECESSARY
  logic:
    operator: AND
    operands:
    - criterion_predicate: OTHER
      description: >-
        Germline monogenic etiology disrupting immune development, regulation,
        or effector function, recognized by IUIS as an inborn error of immunity.
    - criterion_predicate: OTHER
      negated: true
      description: >-
        Not an acquired or secondary immunodeficiency (e.g., HIV/AIDS) and not
        a somatic IEI phenocopy (IUIS Table 10).
  evidence:
  - reference: PMID:41608114
    reference_title: "Human inborn errors of immunity: 2024 update on the classification from the International Union of Immunological Societies Expert Committee."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Inborn errors of immunity (IEIs) are, by definition, caused by damaging
      germline variants in single genes.
    explanation: >-
      The IUIS Expert Committee's definitional statement, which is the positive
      conjunct of these criteria: germline, monogenic. It is also what excludes
      the acquired and somatic-phenocopy conditions named in the negated
      conjunct.
  - reference: PMID:41608114
    reference_title: "Human inborn errors of immunity: 2024 update on the classification from the International Union of Immunological Societies Expert Committee."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      IEIs are currently categorized into 10 tables, with subtables segregating
      groups of disorders into overlapping phenotypes.
    explanation: >-
      Establishes the ten-table structure that the nested sub-groupings mirror
      one-to-one, and therefore that a complete IEI tree needs ten branches
      rather than the four this grouping originally carried.
members:
- member: Immune Dysregulation IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 4 — diseases of immune dysregulation (HLH, ALPS, IPEX/APECED,
      CTLA4/LRBA checkpoint defects, interferonopathies).
- member: Combined Immunodeficiencies IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 1 — combined immunodeficiencies affecting cellular and humoral
      immunity (SCID, CID, Omenn syndrome).
- member: Combined Immunodeficiency with Syndromic Features IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 2 — combined immunodeficiencies with associated or syndromic
      features (DiGeorge, Wiskott-Aldrich, ectodermal dysplasia with
      immunodeficiency, ICF).
- member: Predominantly Antibody Deficiencies IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 3 — predominantly antibody deficiencies (XLA, CVID, selective
      IgA/IgM deficiency, hyper-IgM syndromes).
- member: Congenital Phagocyte Defect IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 5 — congenital defects of phagocyte number or function
      (chronic granulomatous disease, severe congenital neutropenia,
      leukocyte adhesion deficiency).
- member: Intrinsic and Innate Immunity Defect IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 6 — defects in intrinsic and innate immunity, producing
      susceptibility restricted to one pathogen class (Mendelian
      susceptibility to mycobacterial disease, chronic mucocutaneous
      candidiasis, herpesvirus and HPV susceptibility).
- member: Autoinflammatory Disorder IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 7 — autoinflammatory disorders: sterile, antigen-independent
      innate inflammation via inflammasome/IL-1, type I interferon, or
      NF-kappaB dysregulation (CAPS, COPA syndrome, the interferonopathies).
- member: Complement Deficiency IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 8 — complement deficiencies, where the position of the defect
      within the cascade predicts the phenotype (terminal-component defects and
      neisserial infection; early classical-pathway defects and lupus-like
      immune-complex disease).
- member: Bone Marrow Failure IEIs
  member_type: GROUPING
  differentiating_mechanisms:
  - description: >-
      IUIS Table 9 — bone marrow failure: immunodeficiency arising from failure
      of the haematopoietic stem and progenitor compartment through
      genome-maintenance, DNA-replication or telomere-biology lesions.
references:
- reference: PMID:41608114
  title: >-
    Human inborn errors of immunity: 2024 update on the classification from the
    International Union of Immunological Societies Expert Committee.
  findings:
  - statement: >-
      The current (2024) IUIS classification, superseding the June 2022 update
      this grouping was originally written against. Catalogues 508 genes and 17
      phenocopies across ten phenotypic tables, of which 67 monogenic defects
      and 2 phenocopies are new since 2022.
  - statement: >-
      Gives the per-table entity counts the sub-groupings are measured against:
      Table 1, 73 genes; Table 2, 83; Table 3, 48; Table 4, 72 defects; Table 5,
      45 defects; Table 6, 86; Table 7, 69 disorders; Table 8, 36; Table 9, 47;
      Table 10, 17 phenocopies.
- reference: PMID:41608113
  title: >-
    The 2024 update of IUIS phenotypic classification of human inborn errors of
    immunity.
  findings:
  - statement: >-
      The phenotypic companion to the genotypic classification, and the one this
      tree is structurally closest to: it describes 559 IEI and presents the
      classification as decision trees keyed on clinical and immunological
      phenotype rather than on gene. Relevant because the sub-groupings here
      are phenotype-organized (infection vs dysregulation vs autoinflammation)
      and inherit that framing.
discussions:
- discussion_id: iei_tree_coverage_against_iuis_2024
  kind: KNOWLEDGE_GAP
  prompt: >-
    What fraction of the IUIS catalogue does this tree actually represent, and
    is the sample it holds representative or biased?
  rationale: >-
    As of 2026-08-20 the tree reaches 36 curated Disease entries against the
    IUIS 2024 catalogue of 508 genes and 17 phenocopies — roughly 7%. That in
    itself is expected for a mechanism knowledge base that curates in depth
    rather than enumerating. What matters for anyone querying the tree is that
    the coverage is uneven in a way the member counts make visible: Tables 2 and
    4 hold ten and nine entries respectively, while Tables 8 and 9 hold one each
    and Table 10 (phenocopies) is out of scope by design. A query that treats
    this grouping as a sample of IEI will therefore over-weight syndromic
    combined immunodeficiency and immune dysregulation and under-weight
    complement and marrow failure. The gap is curation coverage, not a
    scientific uncertainty, but it should not be inferred from the tree's
    structure alone.
  proposed_experiments:
  - experiment_id: iei_coverage_census_against_iuis_tables
    name: Census KB IEI coverage against the IUIS 2024 per-table counts
    description: >-
      For each of the nine in-scope IUIS tables, compare the count of
      kb/disorders/ entries carrying the matching
      `classifications.iuis_category` against the published per-table entity
      count (Table 1, 73; Table 2, 83; Table 3, 48; Table 4, 72; Table 5, 45;
      Table 6, 86; Table 7, 69; Table 8, 36; Table 9, 47) and publish the
      per-table coverage ratio. This turns the sampling bias above into a
      number a consumer of the grouping can read, and gives curation a
      table-level priority ordering rather than a flat backlog.
- discussion_id: iuis_category_backfill_candidates
  kind: KNOWLEDGE_GAP
  prompt: >-
    Which further kb/disorders/ entries belong in an IUIS table, and what
    evidence fixes the assignment for a disease that is mechanistically an
    obvious IEI but whose table placement has never been curated?
  rationale: >-
    Several entries carry an unmistakable IEI phenotype but no
    `classifications.iuis_category`, so the sub-groupings' NECESSARY criteria
    cannot be evaluated for them and they cannot be listed without asserting a
    table placement the entry does not itself record. Identified 2026-08-20:
    STAT2 Deficiency (a break in type I interferon signal transduction, giving
    the Table 6 signature of susceptibility restricted to viral disease);
    Wiskott-Aldrich Syndrome and Ataxia Telangiectasia (Table 2); and
    Aicardi-Goutieres Syndrome, Familial Cold Autoinflammatory Syndrome and
    Proteasome-Associated Autoinflammatory Syndrome (Table 7). The general
    problem is that table assignment is curated per Disease entry while grouping
    membership is curated on the grouping, so an entry can be mechanistically
    obvious and still structurally invisible to the audit. The gap is
    bookkeeping rather than biology, but leaving it implicit is what let five
    IUIS tables go unrepresented in this tree until 2026-08-20. Hosted on the
    umbrella rather than on any one sub-grouping because the candidates span
    Tables 2, 6 and 7.
  proposed_experiments:
  - experiment_id: backfill_iuis_category_across_candidate_entries
    name: Backfill iuis_category across candidate IEI entries
    description: >-
      For each kb/disorders/ entry with an IEI phenotype and no
      `classifications.iuis_category`, cite the IUIS 2024 classification
      (PMID:41608114) or a disease-specific source that states the table
      placement, add the assignment with a verified snippet, then add the entry
      to the matching sub-grouping. Each needs its own sourced assignment, which
      is why the backfill was not done as part of the pass that surfaced it.
- discussion_id: iei_phenocopy_scope_boundary
  kind: KNOWLEDGE_GAP
  prompt: >-
    Should IUIS Table 10 phenocopies be modelled somewhere in dismech, given
    that they are deliberately excluded from this grouping?
  rationale: >-
    This grouping excludes IUIS Table 10 because its members are not germline
    monogenic disorders and so fail the NECESSARY criteria — 8 of the 17 are
    caused by somatic mutations and 9 by neutralizing anti-cytokine
    autoantibodies. The exclusion is correct for this grouping but leaves the
    concepts homeless: an anti-IFN-gamma autoantibody syndrome is
    mechanistically an immunodeficiency, is diagnosed and managed in the same
    clinics, and phenocopies a curated germline entry closely enough that the
    two are differential diagnoses of each other. Whether dismech should carry
    them as ordinary Disease entries outside this tree, as a sibling
    non-IEI grouping, or not at all is an open scope decision, not something
    the criteria here settle.
  evidence:
  - reference: PMID:41608114
    reference_title: "Human inborn errors of immunity: 2024 update on the classification from the International Union of Immunological Societies Expert Committee."
    supports: SUPPORT
    evidence_source: OTHER
    snippet: >-
      Total number of conditions for Table 10: 17 (8 due to somatic mutations;
      9 due to autoantibodies).
    explanation: >-
      Quantifies the excluded set and confirms its two non-germline mechanisms,
      which is why Table 10 fails this grouping's NECESSARY criteria rather
      than merely being uncurated.
notes: >-
  Nine IUIS-table sub-groupings, one per in-scope table; Table 10 (phenocopies)
  is excluded by the membership criteria rather than uncurated — see the
  `iei_phenocopy_scope_boundary` discussion. Per-disease IUIS assignment belongs
  in `classifications.iuis_category` on each Disease entry, and each
  sub-grouping's NECESSARY criterion is a `HAS_CLASSIFICATION` leaf reading that
  slot, so membership is machine-auditable via `just check-groupings` rather
  than aspirational.

  Expanded 2026-08-20 from four sub-groupings to nine. Tables 5-9 (phagocyte
  defects, intrinsic and innate immunity, autoinflammatory disorders,
  complement deficiencies, bone marrow failure) already had curated members
  carrying the right `iuis_category` but no sub-grouping to hang from, so eight
  IEI entries were unreachable from this umbrella. Thirteen further entries were
  added to the four existing sub-groupings in the same pass.

  Mechanism modules for IEI pathway families remain a follow-up before
  NECESSARY_AND_SUFFICIENT CONFORMS_TO_MODULE criteria can be used. MONDO
  descendant coverage and mapping consistency:
  `uv run python scripts/grouping_mondo_gaps.py` (see research/grouping_mondo_gaps.md).