Curation of gene–disease relationships in primary antibody deficiencies using the ClinGen validation framework

Abstract

Background:

The Clinical Genome Resource (ClinGen) is an international collaborative effort among scientists and clinicians, diagnostic and research laboratories, and the patient community. Using a standardized framework, ClinGen has established guidelines to classify gene–disease relationships as definitive, strong, moderate, and limited on the basis of available scientific and clinical evidence. When the genetic and functional evidence for a gene–disease relationship has conflicting interpretations or contradictory evidence, they can be disputed or refuted.

Objective:

We assessed genes related to primary antibody deficiencies.

Methods:

The ClinGen Antibody Deficiencies Gene Curation Expert Panel, using the ClinGen framework, classified genes related to primary antibody deficiency that primarily affect B-cell development and/or function, and that account for the largest proportion of inborn errors of immunity or primary immunodeficiencies.

Results:

The expert panel curated a total of 65 genes associated with humoral immune defects to validate 74 gene–disease relationships. Of these, 40 were classified as definitive, 1 as strong, 16 as moderate, 15 as limited, and 2 as disputed. The curation process involved reviewing 490 patient records and 3546 associated human phenotype ontology entries. The 3 most frequently observed terms related to primary antibody deficiency were decreased circulating antibody level, pneumonia, and lymphadenopathy.

Conclusions:

These curations (publicly available at ClinicalGenome.org) represent the first effort to provide a comprehensive genetic and phenotypic revision of genetic disorders affecting humoral immunity, as reviewed and approved by experts in the field.

Currently, more than 500 monogenic immune disorders—termed inborn errors of immunity (IEIs) or primary immunodeficiencies (PIDs)—have been described.^1,2 These diseases are classified and updated by the International Union of Immunological Societies (IUIS) Expert Committee on Inborn Errors of Immunity approximately every 2 years.³ Gene defects causing quantitative or functional B-cell defects as the predominant clinical or immunologic feature are classified as primary antibody deficiencies (PADs).³ Moreover, some IEIs are associated with complex phenotypes or involvement of other components of the immune system beyond the B-cell compartment. Some of these IEIs have been linked to a diagnosis of common variable immunodeficiency (CVID). Currently, CVID acts as a catchall term used to categorize patients with variable B-cell defects, both quantitative and qualitative. To date, there are several specific genetic defects associated with CVID, which enables their reclassification according to genotype instead of a generic diagnosis of CVID.^4,5

The Clinical Genome Resource (ClinGen, ClinicalGenome.org) is a National Institutes of Health–funded initiative that aims to build an authoritative central resource for defining the clinical relevance of genes and variants in precision medicine and research.^6,7 As part of ClinGen, the Immunology Clinical Domain Working Group (Immunology CDWG, www.clinicalgenome.org/working-groups/clinical-domain/immunology/#heading_membership) was established in October 2019.⁸ The Immunology CDWG comprises domain experts, clinicians and scientists, genetic counselors, and biocurators from international academic medical centers, research institutions, hospitals, diagnostic laboratories, and industry. The ClinGen Antibody Deficiencies Gene Curation Expert Panel (AD-GCEP, clinicalgenome.org/affiliation/40080/) was formed in October 2020 under the auspices of the Immunology CDWG with the objective to curate genes listed under “Table III: Predominantly Antibody Deficiencies” of the 2020 and 2022 IUIS classification.³ Moreover, additional genes with relevant humoral immunodeficiency phenotypes were included in its mandate, as determined by the AD-GCEP leadership.

Table III.

Harmonization of disease nomenclature for PAD

Gene	MONDO term (ID)	OMIM phenotype term	AD-GCEP–suggested name	Synonyms
ADA2	Deficiency of adenosine deaminase 2 (0100317)	Sneddon syndrome; vasculitis, autoinflammation, immunodeficiency, and hematologic defects syndrome	Adenosine deaminase 2 deficiency	DADA2
AICDA	Hyper-IgM syndrome type 2 (0011528)	Immunodeficiency with hyper-IgM, type 2	AICDA-associated autosomal-recessive hyper-IgM syndrome	AR-AID–hyper-IgM syndrome
AICDA	Hyper-IgM syndrome type 2 (0011528)	Immunodeficiency with hyper-IgM, type 2	AICDA-associated dominant-negative hyper-IgM syndrome	AD-AID–hyper-IgM syndrome, AID-dominant defect
ARHGEF1	Immunodeficiency 62 (0032763)	Immunodeficiency 62	ARHGEF1-related immunodeficiency
ATM	Ataxia telangiectasia (0008840)	Ataxia-telangiectasia	Ataxia telangiectasia
ATP6AP1	Congenital disorder of glycosylation type II (0005501)	Immunodeficiency 47	ATP6AP1-associated congenital disorder of glycosylation
BACH2	Immunodeficiency 60 (0032723)	Immunodeficiency 60 and autoimmunity	BRIDA syndrome	BACH2-related deficiency
BLNK	Agammaglobulinemia 4, autosomal recessive (0013289)	Agammaglobulinemia 4	BLNK-related immunodeficiency	BLNK agammaglobulinemia
BRWD1	Agammaglobulinemia (0015977)	Ciliary dyskinesia, primary, 51	BRWD1-related immunodeficiency	BRWD1 hypogammaglobulinemia
BTK	Isolated growth hormone deficiency type III (0010615)	Isolated growth hormone deficiency, type III, with agammaglobulinemia	Isolated growth hormone deficiency type III
BTK	Bruton-type agammaglobulinemia (0010421)	Agammaglobulinemia, X-linked 1	BTK-related agammaglobulinemia	X-linked agammaglobulinemia, XLA
CARD11	Immunodeficiency 11b with atopic dermatitis (0054697)	Immunodeficiency 11B with atopic dermatitis	CARD11-related immunodeficiency with atopic dermatitis
CARD11	Severe combined immunodeficiency due to CARD11 deficiency (0014081)	Immunodeficiency 11A	CARD11-related combined immunodeficiency
CARD11	BENTA disease (0014645)	B-cell expansion with NFKB and T-cell anergy	BENTA syndrome
CD19	Immunodeficiency, common variable, 3 (0013283)	Immunodeficiency, common variable, 3	CD19-related immunodeficiency
CD40	Hyper-IgM syndrome type 3 (0011735)	Immunodeficiency with hyper-IgM, type 3	CD40-related immunodeficiency	CD40–hyper-IgM syndrome
CD40LG	Hyper-IgM syndrome type 1 (0010626)	Immunodeficiency, X-linked, with hyper-IgM	CD40L-related immunodeficiency	X-linked hyper-IgM syndrome; CD40L–hyper-IgM syndrome
CD79A	Agammaglobulinemia 3, autosomal recessive (0013288)	Agammaglobulinemia 3	Igα-related immunodeficiency	Igα agammaglobulinemia
CD79B	Agammaglobulinemia 6, autosomal recessive (0012987)	Agammaglobulinemia 6	Igβ-related immunodeficiency	Igβ agammaglobulinemia
CD81	Immunodeficiency, common variable, 6 (0013286)	Immunodeficiency, common variable, 6	CD81-related deficiency (CVID)
CR2	Immunodeficiency, common variable, 7 (0013862)	Immunodeficiency, common variable, 7	CD21-related immunodeficiency
CTLA4	Autoimmune lymphoproliferative syndrome due to CTLA4 haploinsufficiency (0014493)	Hashimoto thyroiditis; Susceptibility to Systemic lupus erythematosus. Immune dysregulation with autoimmunity, immunodeficiency, and lymphoproliferation	CTLA4 deficiency	IDAIL, CHAI, CTLA-4 haploinsufficiency with autoimmune infiltration
CTNNBL1	Common variable immunodeficiency (0015517)	Immunodeficiency 99 with hypogammaglobulinemia and autoimmune cytopenias	CTNNBL1-related immunodeficiency
CXCR4	WHIM syndrome (0023880)	Myelokathexis, isolated; WHIM syndrome 1	WHIM syndrome	Warts, hypogammaglobulinemia, immunodeficiency with myelokathexis
FNIP1	FNIP1-associated syndrome (0100432)	Immunodeficiency 93 and hypertrophic cardiomyopathy	FNIP1-related immunodeficiency	B-cell deficiency due to FNIP1
GATA2	GATA2 deficiency with susceptibility to MDS/AML (0042982)	Emberger syndrome; Immunodeficiency 21	GATA2 deficiency	GATA2 haploinsufficiency, DCML deficiency, MonoMAC syndrome
ICOS	Common variable immunodeficiency (0015517)	Immunodeficiency, common variable, 1	ICOS-related immunodeficiency
ICOSLG	Combined immunodeficiency (0015131)	Immunodeficiency 119	ICSOL-related immunodeficiency	Combined immunodeficiency due to ICOSL
IGHM	Autosomal recessive agammaglobulinemia 1 (0020729)	Agammaglobulinemia 1	Igμ heavy chain–related immunodeficiency	Igμ agammaglobulinemia
IGKC	Recurrent infections associated with rare immunoglobulin isotypes deficiency (0013576)	Kappa light chain deficiency	Igκ light chain–related immunodeficiency	Igκ LC agammaglobulinemia
IGLL1	Agammaglobulinemia 2, autosomal recessive (0013287)	Agammaglobulinemia 2	Immunoglobulin light chain–related immunodeficiency	Immunoglobulin LL1 agammaglobulinemia
IKZF1	Pancytopenia due to IKZF1 mutations (0014810)	Immunodeficiency, common variable, 13	Ikaros deficiency
IKZF1	Autoimmune disease (0007179)	No term associated	Ikaros-related combined immunodeficiency	Ikaros gain of function with combined immunodeficiency
IKZF2	HELIOS deficiency (0800139)	No term associated	Helios-related immunodeficiency
IKZF3	Immunodeficiency 84 (0030333)	Immunodeficiency 84	Aiolos-related immunodeficiency
IL21	Common variable immunodeficiency (0015517)	Immunodeficiency, common variable, 11	IL-21–related immunodeficiency
IL21R	Immunodeficiency disease (0021094)	Immunodeficiency 56	IL-21R–associated immunodeficiency
INO80	Immunodeficiency, common variable, 1 (0011864)	No term associated	INO80-related immunodeficiency
IRF2BP2	Immunodeficiency, common variable, 14 (0054691)	Immunodeficiency, common variable, 14	IRF2BP2-related immunodeficiency
IRF4	Combined immunodeficiency (0015131)	Variation in skin/hair/eye pigmentation, 8	IRF4-related immunodeficiency	IRF4-related combined immunodeficiency
LRBA	Combined immunodeficiency due to LRBA deficiency (0013863)	Immunodeficiency, common variable, 8, with autoimmunity	LRBA deficiency	LATAIE
LRRC8A	Agammaglobulinemia (0015977)	Agammaglobulinemia 5	LRRC8-related immunodeficiency	LRRC8 agammaglobulinemia
MAP3K14	NIK deficiency (0018642)	Immunodeficiency 112	NIK-related immunodeficiency
MOGS	MOGS-congenital disorder of glycosylation (0011629)	Congenital disorder of glycosylation, type IIb	MOGS-congenital disorder of glycosylation
MS4A1	Immunodeficiency, common variable, 5 (0013285	Immunodeficiency, common variable, 5	CD20-related immunodeficiency
NFKB1	Immunodeficiency, common variable, 12 (0014697)	Immunodeficiency, common variable, 12	NFKB1-related immunodeficiency
NFKB2	Immunodeficiency, common variable, 10 (0014260)	Immunodeficiency, common variable, 10	NFKB2-related immunodeficiency	CVID; DAVID syndrome
PAX5	PAX5-related B lymphopenia and autism spectrum disorder (0100299)	Leukemia, acute lymphoblastic, susceptibility to, 3	PAX5-related immunodeficiency with autism spectrum disorder	PAX5-related B-cell deficiency
PIK3CD	Immunodeficiency 14 (0014222)	Immunodeficiency 14A, autosomal dominant	APDS type 1	APDS type 1; PASLI
PIK3CD	Immunodeficiency 14b, autosomal recessive (0023655)	Immunodeficiency 14B, autosomal recessive	PIK3CD (p110δ) deficiency
PIK3CG	Immunodeficiency 97 with autoinflammation (0030717)	Immunodeficiency 97 with autoinflammation	PIK3CG-related defects	PIK3CG loss-of-function defect
PIK3R1	Immunodeficiency 36 (0014453)	Immunodeficiency 36	APDS type 2	APDS type 2
PIK3R1	Agammaglobulinemia 7, autosomal recessive (0014083)	Agammaglobulinemia 7, autosomal recessive	p85a-related immunodeficiency	p85a agammaglobulinemia
PMS2	Mismatch repair cancer syndrome 1 (0010159)	Mismatch repair cancer syndrome 4	PMS2-related immunodeficiency
POU2AF1	Agammaglobulinemia (0015977)	No term associated	Bob-1–related immunodeficiency	Bob-1 agammaglobulinemia
PTEN	PTEN hamartoma tumor syndrome (0017623)	Cowden syndrome 1	PTEN hamartoma tumor syndrome
RAC2	Immunodeficiency 73b with defective neutrophil chemotaxis and lymphopenia (0033554)	Immunodeficiency 73B with defective neutrophil chemotaxis and lymphopenia	RAC2-related combined immunodeficiency	RAC2 gain of function
RAC2	Immunodeficiency 73c with defective neutrophil chemotaxis and hypogammaglobulinemia (0033555)	Immunodeficiency 73C with defective neutrophil chemotaxis and hypogammaglobulinemia	RAC2 deficiency	RAC2 hypogammaglobulinemia with impaired neutrophil chemotaxis
RAC2	Neutrophil immunodeficiency syndrome (0011988)	Immunodeficiency 73A with defective neutrophil chemotaxis and leukocytosis	RAC-2 neutrophil dominant deficiency
SASH3	Combined immunodeficiency, X linked (0010730)	Immunodeficiency 102	SASH3-related combined immunodeficiency	X-linked–related SASH3 defect
SEC61A1	SEC61A1 deficiency (0100337)	Severe congenital neutropenia, 11, autosomal dominant	SEC61A1-related immunodeficiency
SERPING1	Hereditary angioedema with C1INH deficiency (0033946)	Partial deficiency of complement component 4	Hereditary angioedema related to C1 inhibitor deficiency (aka HAE)	C1INH deficiency
SH3KBP1	Immunodeficiency 61 (0010296)	Immunodeficiency 61	CIN85-related immunodeficiency
SLC39A7	Agammaglobulinemia (0015977)	Agammaglobulinemia 9, autosomal recessive	ZIP7-related immunodeficiency	ZIP7 agammaglobulinemia
SPI1	Agammaglobulinemia 10, autosomal dominant (0030529)	Agammaglobulinemia 10, autosomal dominant	PU.1-related immunodeficiency	PU.1 agammaglobulinemia
TCF3	Autosomal agammaglobulinemia (0011096)	Agammaglobulinemia 8A, autosomal dominant Agammaglobulinemia 8B, autosomal recessive	TCF3-related immunodeficiency	TCF3 agammaglobulinemia
TNFRSF13B	Immunodeficiency, common variable, 2 (0009413)	Immunodeficiency, common variable, 2	Autosomal-recessive TACI-related immunodeficiency (CVID); homozygous TACI deficiency*
TNFRSF13C	Immunodeficiency, common variable, 4 (0013284)	Immunodeficiency, common variable, 4	BAFF-R–related immunodeficiency (CVID)
TNFSF12	Common variable immunodeficiency (0015517)	No term associated	TWEAK-associated immunodeficiency
TNFSF13	Common variable immunodeficiency (0015517)	No term associated	APRIL-related immunodeficiency (CVID)
TOP2B	B-cell immunodeficiency, distal limb anomalies, and urogenital malformations (0012243)	B-cell immunodeficiency, distal limb anomalies, and urogenital malformations	TOP2B-related immunodeficiency	Hoffman syndrome
TRAF3	TRAF3 haploinsufficiency (0100513)	Susceptibility to encephalopathy, acute, infection-induced (herpesspecific), 5	TRAF3 deficiency	TRAF3 haploinsufficiency
TRNT1	Congenital sideroblastic anemia–B-cell immunodeficiency–periodic fever–developmental delay syndrome (0014487)	Sideroblastic anemia with B-cell immunodeficiency, periodic fevers, and developmental delay	Sideroblastic anemia with B-cell immunodeficiency, periodic fevers, and developmental delay	SIFD, TRNT1-related deficiency
UNG	Hyper-IgM syndrome type 5 (0011971)	Immunodeficiency with hyper-IgM, type 5	UNG-related immunodeficiency	UNG–hyper-IgM syndrome

APRIL, Aproliferation-inducing ligand; BRIDA, Bach2-related immunodeficiency and autoimmunity; DAVID, deficient anterior pituitary with variable immune deficiency; LATAIE, LRBA deficiency with autoantibodies, regulatory T-cell defects, autoimmune infiltration, and enteropathy; MonoMAC, monocytopenia and mycobacterial infection; PASLI, p110δ-activating variant related to T-cell senescence, lymphadenopathy, and immunodeficiency; TACI, transmembrane activator and CAML interactor; TWEAK, TNF-like weak inducer of apoptosis; WHIM, warts, hypogammaglobulinemia, infections, and myelokathexis.

^*Heterozygous TNFRSF13B (TACI) variants have not been formally curated because these are considered risk variants for CVID, due to insufficient evidence as independently associated with disease causality.

Here, we describe the approach to curation of genes associated with PADs using a standardized framework, which included efforts to harmonize disease nomenclature with other public databases such as Online Mendelian Inheritance in Man (OMIM; www.omim.org) and Mondo Disease Ontology (MONDO; mondo.monarchinitiative.org), and identification of Human Phenotype Ontology (HPO; hpo.jax.org) terms related to humoral immune defects. We curated and classified 65 genes related to PADs, resulting in the classification of 74 gene–disease relationships, in a process that included phenotypic description of 490 patients using 3546 HPO terms. The utilization of the ClinGen gene curation framework offers a unique opportunity for applying evidence-based guidelines to determine gene–disease relationships.^6,9

METHODS

ClinGen framework

ClinGen has developed the infrastructure and documentation to provide harmonized and standardized gene and variant curation procedures. Experts in the field were invited to review the curations that follow this framework. The process utilizes gene–disease validity standard operating procedures to perform the gene–disease curations.¹⁰ Version 10.1 of this standard operating procedure is publicly available on the ClinGen website and is summarized in Fig 1. ClinGen curations are updated periodically. The most current information is maintained at ClinicalGenome.org.

FIG 1.

ClinGen gene–disease curation framework. ClinGen gene–disease validity framework started with selection of genes within scope of GCEP and precuration. After this, curators assessed and scored all genetic and experimental evidence reported for each gene–disease entity. Classification was reviewed by expert members of GCEP and published on ClinGen’s website (ClinicalGenome.org) after approval. Recuration for genes not already recurated is planned in next 3 years, or sooner if contradictory evidence emerges (clinical-genome.org/site/assets/files/2164/clingen_standard_gene-disease_validity_recuration_procedures_v1.pdf).

Lumping and splitting

Several PAD genes are reported in relation to multiple disease entities in public databases or literature, and these gene–disease relationships may differ in the molecular mechanisms of disease, phenotypic expression, or mode of inheritance. These criteria determine whether a gene–disease relationship needs to be curated independently (ie, split) or together (ie, lumped) on the bases of assertions in OMIM, MONDO, and literature reports as well as molecular mechanism; whether the phenotype is variable among the assertions; inheritance patterns (autosomal recessive [AR], autosomal dominant [AD], X-linked recessive or dominant, semidominant); and variable functional effects (gain of function, loss of function, haploinsufficiency, dominant negative). The compilation of this evidence for the lumping and splitting decision, or precuration, a documented process in the ClinGen Framework.¹¹ Curators completed their precuration within the ClinGen Gene Tracker program.¹²

Gene curation

Gene curation involves a scoring system that considers published genetic and experimental evidence as outlined in the ClinGen Gene–Disease Validity Standard Operating Procedures v11 (www.clinicalgenome.org/docs/gene-disease-validity-standard-operating-procedures-version-11/). Within genetic evidence, a maximum of 12 points could be awarded for the data. The genetic evidence evaluates factors such as case-level data, case–control data, and type of variants—for example, null variants (scoring 11.5 points), other variant types such as missense (scoring 10.1 points), variant-level functional data (evidence generated in an isogenic system and/or overexpression, scoring 10.5 points), and de novo status (scoring 10.5 points). The cumulative score from published cases established the total score for the genetic evidence. While a single patient case report or single related kindreds can be used within the curation framework, they would not independently provide a classification status of definitive for a particular gene, even with biochemical and functional validation, and would therefore remain as a limited classification without additional patient findings. However, the value of single patient reports with substantial functional, biochemical, and genetic evidence cannot be discounted in the classification of IEIs/PIDs as a result of the nature of rare and ultrarare diseases (see Fig E1, A, in the Online Repository at www.jacionline.org).¹³

Regarding experimental evidence, a maximum of 6 points can be assigned. This evidence includes functional assays to elucidate gene function, as well as evidence supporting the role of the specific gene in a disease or related phenotypic and/or clinical features. Functional evidence categories include biochemical function, protein expression and interaction, functional alteration in patient cells, cell models, and animal models (Fig E1, B).

The sum of the genetic and experimental scores based on published evidence determines the final classification of the gene–disease relationship (Fig 1); if a gene–disease relationship exceeds the maximum scoreable points (12 for genetic evidence and 6 for experimental evidence), then that evidence is included in the curation but does not affect the final classification. If the publicly available evidence includes reports of findings that are replicated over time, then a gene–disease relationship can reach a classification of definitive. A minimum of 3 years and additional independent observations over time of a gene–disease relationship is required to upgrade the classification to definitive status. The chairs and coordinators of the AD-GCEP review the literature for available information, which would determine the appropriate time to initiate a recuration process. In addition, ClinGen offers a newsletter for curated genes that indicates if new reports are publicly available.

HPO terms and principal component analysis

We collected all the HPO terms associated with every patient that was used for genetic evidence. This was a detailed process requiring careful review to generate an HPO database associated with every gene–disease entity for PADs. Once collated, 3546 terms in total were subdivided into 3 categories: clinical manifestations, infections, and laboratory results. A principal component analysis (PCA) graph was built using the frequencies of each HPO term and their association with the genes. The best representation of a given HPO signature was represented by color intensity using the calculated cos² value—that is, the squared cosine value, which represents the quality of a variable’s representation on a specific principal component.

We particularly note that the CD40LG gene was curated by the Severe Combined Immunodeficiency Gene Curation Expert Panel (GCEP), but the data from that curation are included here because the AD-GCEP has curated some of the other genes associated with hyper-IgM defects.

RESULTS

Genes included in scope of AD-GCEP

The AD-GCEP included genes that affect the development and/or function of B cells. The majority of these genes were derived from “Table III: Predominantly Antibody Deficiencies,” with a total of 36 genes.¹⁴ Another 25 genes were included from the 2022 IUIS updated classification.³ Three additional genes were included because publications had identified a related phenotype with humoral immune defects: BRWD1, PAX5, and LRRC8A.^15–17 Therefore, a total of 65 genes were curated, resulting in classification of 74 gene–disease relationships.

Lumping and splitting before curation

Among the 65 genes, 7 were found to be related to more than one disease entity in OMIM, MONDO, or the literature: PIK3R1, PIK3CD, CARD11, IKZF1, RAC2, BTK, and AICDA. Therefore, these genes underwent precuration according to ClinGen’s lumping and splitting criteria (clinicalgenome.org/working-groups/lumping-and-splitting/) (Table I). The AD-GCEP’s complete list of split assertions is provided in Table II.

TABLE I.

Split gene–disease relationships in genes causing PAD

Gene	Mode of inheritance	Disease*	Description (abbreviation)
PIK3R1	AD	SHORT syndrome†
	AD	Immunodeficiency 36	Activated PI3K-δ syndrome 2 (APDS-2)
	AR	Agammaglobulinemia 7, AR	AR agammaglobulinemia due to p85α deficiency
PIK3CD	AD (GOF)	Immunodeficiency 14	Activated PI3K-δ syndrome 1 (APDS-1)
	AR	Immunodeficiency 14b, AR	AR PIK3CD defect/deficiency
CARD11	AD	Immunodeficiency 11b with atopic dermatitis	CARD11 deficiency associated with atopic dermatitis
	AD (GOF)	BENTA disease	BENTA syndrome
	AR	Severe combined immunodeficiency due to CARD11 deficiency	CARD11 deficiency causing combined immunodeficiency (CID)
IKZF1	AD	Pancytopenia due to IKZF1 mutations	Ikaros haploinsufficiency (CVID)
	AD (GOF)	Autoimmune disease	Ikaros gain of function with combined immunodeficiency
RAC2	AD (GOF)	Immunodeficiency 73b with defective neutrophil chemotaxis and lymphopenia	Autosomal-dominant RAC2 gain-of-function defect
	AR	Immunodeficiency 73c with defective neutrophil chemotaxis and hypogammaglobulinemia	RAC2 deficiency with hypogammaglobulinemia
	AD	Neutrophil immunodeficiency syndrome	—
BTK	XL	Isolated growth hormone deficiency type III	—
	XL	Bruton-type agammaglobulinemia	X-linked agammaglobulinemia (XLA)
AICDA	AD	AICDA-associated dominant negative hyper-IgM syndrome	—
	AR	AICDA-associated AR hyper-IgM syndrome	—

GOF, Gain of function; SHORT, short stature, hyperextensibility, ocular depression, Rieger anomaly, and teething delay.

^*MONDO terms associated with split entities.

^†PIK3R1 in relationship with SHORT syndrome was curated by Actionability Working Group. PIK3R1 gene will be curated again in 2025, and AD immunodeficiency will be lumped with SHORT syndrome.

TABLE II.

Classification of gene–disease entities curated by AD-GCEP using ClinGen framework

Gene	ClinGen ID	AD-GCEP disease nomenclature	No. of patients	Mode of inheritance	Genetic evidence	Experimental evidence	Total score	Classification
ADA2	4053	Adenosine deaminase 2 deficiency	8	AR	>12	>6	18	Definitive
AICDA	4080	AICDA-associated autosomal-recessive hyper-IgM syndrome	13	AR	>12	6	18	Definitive
AICDA	4081	AICDA-associated dominant-negative hyper-IgM syndrome	7	AD	7	5.5	12.5	Definitive
ATM *	4205	Ataxia telangiectasia	6	AR	>12	>6	18	Definitive
ATP6AP1 *	4212	ATP6AP1-congenital disorder of glycosylation	10	AR	9.35	2	11.35	Definitive
BLNK	4263	BLNK-related immunodeficiency	5	AR	7.5	4.5	12	Definitive
BTK	4297	BTK-related agammaglobulinemia	25	XL	>12	6	18	Definitive
CARD11	4343	CARD11-related immunodeficiency with atopic dermatitis	21	AD	10.5	5	15.5	Definitive
CARD11	4342	CARD11-related combined immunodeficiency	4	AR	9.2	6	15.2	Definitive
CARD11	4341	BENTA syndrome	12	AD	9	4	13	Definitive
CD19	4372	CD19-related immunodeficiency	7	AR	11.2	>6	17.2	Definitive
CD40	4378	CD40-related immunodeficiency	8	AR	7.6	4.5	12.1	Definitive
CD40LG †	4379	CD40L-related immunodeficiency	4	XL	>12	>6	18	Definitive
CD79A	4380	Igα-related immunodeficiency	7	AR	>12	6	18	Definitive
CD79B	4381	Igβ-related immunodeficiency	3	AR	6	6	12	Definitive
CR2	4573	CD21-related immunodeficiency	3	AR	9	6	15	Definitive
CTLA4	4595	CTLA4 deficiency	16	AD	>12	>6	18	Definitive
CXCR4	4611	WHIM syndrome	9	AD	>12	6	18	Definitive
FNIP1	4872	FNIP1-related immunodeficiency	5	AR	10	5.25	15.25	Definitive
GATA2 *	4924	GATA2 deficiency	18	AD	>12	5	17	Definitive
ICOS	5107	ICOS-related immunodeficiency	10	AR	12	>6	18	Definitive
IGHM	5119	Igm heavy chain–related immunodeficiency	8	AR	>12	1.5	13.5	Definitive
IKZF1	5127	Ikaros deficiency	21	AD	>12	>6	18	Definitive
IL21R	5134	IL-21R-associated immunodeficiency	3	AR	9	4	13	Definitive
IRF4	8358	IRF4-related immunodeficiency	9	AD	6.5	>6	12.5	Definitive‡
LRBA	5299	LRBA deficiency	5	AR	>12	5.5	17.5	Definitive
MOGS *	5405	MOGS–congenital disorder of glycosylation	9	AR	10.25	6	16.25	Definitive
NFKB1	5634	NF-κB1–related immunodeficiency	10	AD	>12	6	18	Definitive
NFKB2	5635	NF-κB2–related immunodeficiency	23	AD	>12	5.5	17.5	Definitive
PIK3CD	5804	Activated p110δ syndrome (APDS) type 1	23	AD	>12	>6	18	Definitive
PIK3CD	5803	PIK3CD (p110δ) deficiency	3	AR	9	3.5	12.5	Definitive
PIK3R1		Activated p110δ syndrome (APDS) type 2	16	AD	>12	3.5	15.5	Definitive
PMS2 *	5842	PMS2-related immunodeficiency	6	AR	>12	>6	18	Definitive
PTEN *	8174	PTEN hamartoma tumor syndrome	3	AD	>12	6	18	Definitive
SASH3	6055	SASH3-related combined immunodeficiency	3	XL	6.25	>6	12.25	Definitive
SERPING1	6113	Hereditary angioedema related to C1 inhibitor deficiency (aka HAE)	13	AD	>12	>6	18	Definitive
SPI1	6265	PU.1-related immunodeficiency	6	AD	10.5	3	13.5	Definitive
TCF3	6337	TCF3-related immunodeficiency	5	Semidominant	9	4.5	13.5	Definitive
TNFRSF13B *	6395	Autosomal-recessive TACI-related immunodeficiency (CVID); homozygous TACI deficiency	13	AR	>12	5	17	Definitive
TRNT1	6449	Sideroblastic anemia with B-cell immunodeficiency, periodic fevers, and developmental delay	5	AR	>12	3	15	Definitive
RAC2	5951	RAC2-related combined immunodeficiency	5	AD	5.5	6	11.5	Strong
BACH2	4230	Bach2-related immunodeficiency and autoimmunity (aka BRIDA)	3	AD	2	5	7	Moderate
ICOSLG	5108	ICSOL-related immunodeficiency	2	AR	2	5	7	Moderate
IGLL1	5122	Immunoglobulin light chain–related immunodeficiency	5	AR	6	2	8	Moderate
IKZF1	5126	Ikaros-related combined immunodeficiency	4	AD	3	>6	9	Moderate
IKZF2	5128	Helios-related immunodeficiency	7	AD	3.95	>6	9.95	Moderate
IKZF3	5129	Aiolos-related immunodeficiency	2	AD	1	>6	7	Moderate
MAP3K14	5333	NIK-related immunodeficiency	3	AR	5	>6	11	Moderate
PAX5	5734	PAX5-related immunodeficiency with autism spectrum disorder	1	AR	3	>6	9	Moderate
PIK3CG	5805	PIK3CG-related defects	2	AR	2.2	>6	8.2	Moderate
RAC2	5950	RAC2 deficiency	2	AR	3.2	6	9.2	Moderate
RAC2	5952	RAC-2 neutrophil-dominant deficiency	2	AD	2	>6	8	Moderate
SEC61A1	6100	SEC61A1-related immunodeficiency	6	AD	8.5	2	10.5	Moderate
SLC39A7	6186	ZIP7-related immunodeficiency	5	AR	4	4	8	Moderate
TOP2B	6415	TOP2B-related immunodeficiency	6	AD	3.75	>6	9.75	Moderate
TRAF3	6431	TRAF3 deficiency	7	AD	8.5	3	11.5	Moderate
ARHGEF1	4164	ARHGEF1-related immunodeficiency	1	XLR	3	>6	9	Limited§
BRWD1	4290	BRWD1-related immunodeficiency	3	AD	1	3	4	Limited
CD81	4382	CD81-related deficiency (CVID)	1	AD	3	6	9	Limited§
CTNNBL1	4601	CTNNBL1-related immunodeficiency	1	AR	3	2.5	5.5	Limited
IGKC	5121	Igκ light chain–related immunodeficiency	5	AR	0.7	0	0.7	Limited
IL21	5133	IL-21–related immunodeficiency	1	AR	1	3	4	Limited
IRF2BP2	5152	IRF2BP2-related immunodeficiency	1	AD	0.5	1	1.5	Limited
LRRC8A	8345	LRRC8-related immunodeficiency	1	AR	2	4	6	Limited
MS4A1	5426	CD20-related immunodeficiency	1	AD	3	2	5	Limited
PIK3R1	5807	p85a-related immunodeficiency	10	AR	3	2.5	5.5	Limited
POU2AF1	5865	Bob-1–related immunodeficiency	1	AR	2.5	>6	8.5	Limited§
SH3KBP1	6131	CIN85-related immunodeficiency	1	AD	1.5	4	5.5	Limited
TNFRSF13C	6396	BAFF-R–related immunodeficiency (CVID)	6	AD	2	>6	8	Limited§
TNFSF12	6398	TWEAK-associated immunodeficiency	2	XL	1.5	2.5	4	Limited
TNFSF13	6399	APRIL-related immunodeficiency (CVID)	1	AR	2	4.25	6.25	Limited
BTK	004296	Isolated growth hormone deficiency type III	3	XL	2.1	1.5	3.6	Disputed
INO80	005145	INO80-related immunodeficiency	2	AR	0	0.5	0.5	Disputed

Genes are organized according to ClinGen classification.

APRIL, Aproliferation-inducing ligand; NF-κB, nuclear factor kappa–light-chain enhancer of activated B cells; TACI, transmembrane activator and CAML interactor; TWEAK, TNF-like weak inducer of apoptosis; WHIM, warts, hypogammaglobulinemia, infections, and myelokathexis; XLR, XL recessive.

^*Gene–disease association where AD-GCEP served as secondary contributor.

^†Curated by Severe Combined Immunodeficiency GCEP (search.clinicalgenome.org/kb/gene-validity/CGGV:assertion_4184ad10-d95f-4027-b75d-e4e11127febf-2021-03-22T154242.503Z). These are reported here with our AD-GCEP curations to maintain all genes associated with hyper-IgM syndrome together.

^‡Classification obtained under recuration process.

^§Classification changed from moderate to limited, given paucity of cases reported in literature.

Gene curation

The final gene–disease classifications were based on the total score, including the available genetic and experimental evidence. Among the 74 curations, 40 were classified as definitive, 1 as strong, 16 as moderate, 15 as limited, and 2 as disputed (Fig 2, A). The most common mode of inheritance was AR (39 curations), followed by AD (28 curations), X-linked (XL; 6 curations), and semidominant (1 curation) (Fig 2, B). Semidominant inheritance, also called incompletely dominant inheritance, refers to an inheritance pattern where the phenotypic severity is related to the number of alleles affected (eg, monoallelic vs biallelic). Among the 65 genes included in our discussion here, only TCF3 presents with a semidominant mode of inheritance.^18–20

FIG 2.

AD-GCEP gene curation. (A) Gene classification based on genetic and experimental evidence: definitive, n = 40; strong, n = 1; moderate, n = 16; limited, n = 15; disputed, n = 2. (B) Mode of inheritance of 74 gene–diseases relationship curated. SD, Semidominant.

The evidence for all curations is provided in Table II and is publicly available on the ClinGen website. We next turn to representative curations for gene–disease relationship classifications: for definitive classifications, BTK and PIK3CD; moderate, TRAF3; and limited, POU2AF1.

Definitive classification: BTK—Bruton-type/XL agammaglobulinemia

Bruton agammaglobulinemia was first described in 1952 as a clinical entity and was subsequently termed XL agammaglobulinemia (XLA).²¹ Four decades later, the gene defect causing this disease was identified as the Bruton tyrosine kinase gene (BTK encoding the protein BTK).^22,23 BTK is a tyrosine kinase that plays a critical role during the early stages of B-cell development in bone marrow. Therefore, the absence of functional BTK causes a failure of early B-cell development, resulting in a complete lack of mature B cells in the periphery and loss of antibody production, leading to recurrent sinopulmonary and other infections in these patients.

Genetic evidence (12 points).

Over 900 unique variants have been reported in humans in the BTKbase,²⁴ including missense, nonsense, frameshift, splicing, and small or large insertions or deletions, known as indels. Evidence supporting this gene–disease relationship includes case-level data, segregation data, and experimental data. Twenty-three unique variants from 23 probands in 4 publications were included in this curation.^23,25–27 Variants in this gene segregated with disease in 8 additional family members.^25,27 Further evidence was available in the literature, but because the maximum score for genetic evidence was reached, these were not included in the final curation.

Experimental evidence (6 points).

BTK plays an essential role in B-cell development, although expression of this protein is not limited to B cells. Indeed, monocytes and platelets also express BTK.²³ As previously stated, loss of BTK protein results in impaired B-cell development and an absence of mature B cells in the periphery with absent or impaired immunoglobulin production. Mouse models of Btk deficiency (eg, Xid/Btk knockout mice) share some features of the human disease (B-cell lymphopenia, hypogammaglobulinemia) but have a milder immunologic phenotype compared to XLA patients.²⁸ The experimental data provided a score of 6 points, the maximum.

In summary, there is definitive-level evidence to support the relationship between BTK and XLA, with a maximum score of 18 points from genetic and experimental evidence. Further, this evidence has been repeatedly supported by additional data in the research and clinical domains and has stood the test of time.

Definitive classification: PIK3CD—APDS-1

PIK3CD, encoding the p110δ catalytic subunit of phosphatidylinositol 3-kinase, was first reported in relation to AD immunodeficiency 14 in 2006.²⁹ However, at the time of this report, the mechanism of disease pathogenesis was unknown. This disease is now recognized as activated p110δ syndrome 1 (APDS-1) and is characterized by recurrent infections (often sinopulmonary), hypogammaglobulinemia (IgG and/or IgA but often elevated IgM), impaired immune responses after natural infection or vaccination, susceptibility to herpesviruses, splenomegaly, lymphoid nodules on mucosa, abscesses, decreased circulating B cells, autoimmunity (cytopenias), and B-cell lymphoma.

Genetic evidence (12 points).

Variants in PIK3CD have been reported in at least 22 probands in 12 publications. The mechanism of disease is gain-of-function missense variants found across the gene, which cause constitutive basal activation of PI3K p110δ.^30–36 Additional evidence is available in the literature, but the maximum score for genetic evidence was reached with the literature cited above.

Experimental evidence (6 points).

This gene–disease relationship is supported by in vitro functional assays and animal models. The p110δ protein is predominantly expressed in leukocytes.^37,38 Several knock-in mouse models have been reported in the literature that recapitulate the cellular phenotype and provide insights into the function of p110δ in immune processes.^35,39–41 In summary, there is definitive evidence to support the relationship between variants in PIK3CD and APDS-1. This has been repeatedly demonstrated in both the research and clinical diagnostic settings and has been upheld over time.

Moderate classification: TRAF3—TRAF3 haploinsufficiency

TRAF3 deficiency as an AD defect was first reported in 2010 in a patient with herpes simplex encephalitis,⁴² but more recently in 2022, it has been reported as an AD disorder in the context of immune dysregulation with hypergammaglobulinemia and B-cell lymphoproliferation.⁴³ TRAF3 encodes TNF receptor–associated factor 3, which is involved in cell signaling and functions downstream of CD40 to activate the immune response. TRAF3 haploinsufficiency presents as an immune dysregulation syndrome characterized by recurrent bacterial infections, autoimmunity, systemic inflammation, B-cell lymphoproliferation, and hypergammaglobulinemia.⁴³

Genetic evidence (8.5 points).

Five unrelated probands were reported with 4 heterozygous loss-of-function variants resulting in reduced TRAF3 messenger RNA and protein expression.⁴³ Two additional case reports suggested a disease phenotype related to TRAF3 missense variants with a likely dominant-negative mechanism.^42,44

Experimental evidence (3.0 points).

Conditional B-cell–and/or T-cell–specific deletion of TRAF3 in transgenic mice recapitulated many of the human phenotypes, including lymphadenopathy, splenomegaly, hypergammaglobulinemia, susceptibility to infection, and autoimmunity.^45–47

In summary, there is evidence to support the relationship between TRAF3 and AD TRAF3 haploinsufficiency, resulting in a classification of moderate, with a total of 11.5 points.

Limited classification: Mutations in POU2AF1, encoding BOB1/OBF-1/OCA-B, causing AR agammaglobulinemia

POU2AF1, or POU domain class 2–associating factor 1, which encodes a protein known as OBF1 (OCT-binding factor 1), BOB1, or OCAB, was first reported in relation to AR agammaglobulinemia in 2021.⁴⁸

Genetic evidence (2.5 points).

To date, only one patient with a BOB1 deficiency caused by a homozygous POU2AF1 null variant, resulting in a severe B-cell–intrinsic defect, has been reported.⁴⁸ B-cell numbers were normal but had disrupted B-cell receptor signaling and plasmablast development, and immunoglobulin secretion appeared to be abrogated in this patient’s cells.

Experimental evidence (6 points).

The highly B lymphocyte–specific pattern of POU2AF1 expression, together with its selective transcriptional regulatory effect on immunoglobulin promoters, strongly indicates that POU2AF1 is a critical determinant of immunoglobulin transcription in lymphoid cells.⁴⁹ This function is deficient in patient cells but was rescued by lentiviral reconstitution.⁴⁸ The immunologic manifestations in the patient showed some overlap with the Pou2af1 gene–targeted mice.^50–53 Patient T cells were normal in terms of distribution and activation, except for a reduction in circulating T follicular helper cells,⁴⁸ which is likely the result of impaired B-cell differentiation and function.²⁹

In summary, there is limited evidence to support the gene–disease relationship between POU2AF1 and OBF1/BOB1/OCAB deficiency leading to AR agammaglobulinemia. In the future, identification of additional patients could allow reclassification, and recuration of this gene will be undertaken in 2025.

HPO terminology for classification of clinical features of PAD

The AD-GCEP curated 74 gene–disease relationships, resulting from the compilation of 490 published patient case reports. All the HPO terms were extracted from all the patient data used for genetic evidence from each curation. The clinical description was adapted and matched with the HPO terms, reaching a total of 3546 HPO terms for PAD; 961 of these HPO terms were reported at least once. The HPO terms were divided into 3 categories: clinical manifestations (49.94%), laboratory test results (31.44%), and infections (18.61%) (Fig 3, A). The top 10 HPO terms for each category are depicted through heat maps that indicate the frequency of the HPO terms across the curated genes (Fig 3, B). Infections are a common feature in patients with PAD, most commonly pneumonia, with a total 81 patients having this clinical feature, and was observed in 31 gene–disease relationships. Additionally, the most representative laboratory test result for this cohort was represented by the HPO term decreased circulating antibody level, which was identified in 126 patients and related to 38 gene–disease entities. Finally, the term lymphadenopathy was noted in 60 patients and was related to 16 gene–disease relationships. PCA demonstrated homogeneity between the HPO terms associated with the gene–disease relationships curated by the AD-GCEP. Nevertheless, when subdividing the HPO terms into the categories of infections, laboratory results, and clinical manifestations, PCA showed significant differences in relation to the HPO terms associated with the following genes: PIK3CD, CARD11, NFKB2, BTK, SPI1, CTLA4, and NFKB1 (Fig 4). This compilation of HPO terms associated with PAD represents the first curated collection, with future potential in machine learning and artificial intelligence algorithms. The complete list of HPO terms for a specific gene can be found on the ClinGen website.

FIG 3.

Most common HPO terms associated with PAD. (A) Frequency (%) of HPO terms associated with clinical manifestations, laboratory results, and infections from 3546 HPO terms in patients with PAD. (B) Single-gradient heat map of top 10 HPO terms found with genes associated with PAD. Color gradient represents higher frequencies of patients presenting that HPO term; number inside each cell is number of patients manifesting HPO term.

FIG 4.

PCA of HPO terms associated with PAD. PCA analysis represents distribution of HPO and genes associated with antibody deficiencies. Only top 10 HPO terms for each category are provided; infections, laboratory results, and clinical manifestations are represented. Color code indicates quality of representation, from strong (red) to weak (green).

Nomenclature in MONDO and OMIM and usage in clinical immunology

Harmonization of disease names was essential during this curation effort. Before attempting curation, information on disease assertions for each gene was gathered from the OMIM and MONDO databases. However, many of these assertions followed the naming pattern of the specific database, which was based on the time they were first reported in the literature or based on the mode of inheritance. For example, variants in PIK3CD are described in OMIM to cause immunodeficiency 36, a generic term that does not represent the phenotype or the disease in the clinical immunology field. Additionally, several genes did not have any associated OMIM or MONDO terms. This nomenclature was not intuitive and did not reflect the gene–disease relationship, so the AD-GCEP compiled all assertions found in MONDO and OMIM for the genes under consideration and harmonized disease terminology according to clinical phenotype and usage within the clinical immunology community (Table III). This standardized nomenclature will be offered to these databases for consideration of updating their existing terms to provide better clinical recognition and encourage universal adoption of a common terminology.

DISCUSSION

The classification of genes associated with PAD using the standardized ClinGen Framework represents a valuable contribution to the field of clinical immunology: it provides an expert, consensus-based, data-driven assignment of genes to clinical conditions. The inclusion of these data on the ClinGen website in real time will ensure their quick and rapid dissemination to the clinical immunology community.

Among the 74 gene–disease associations curated by AD-GCEP, 40 were defined as definitive, with an abundance of both genetic and experimental data. The periodic recuration that the AD-GCEP followed within the ClinGen Framework had allowed the recent reclassification from strong to definitive of genes including IRF4, SPI1, SASH3, and FNIP1. The functional validation of these genes, including genetic rescue of the phenotype in patient cells or nonhuman model, was essential to establishing gene–disease causality. For SASH3, studies in the Jurkat cell line model confirmed the rescue of the functional defect.⁵⁴ For SPI1, SPI1-deficient cells were generated from pro–B-cell lines and hematopoietic pluripotent stem cells by gene editing (clustered regularly interspaced short palindromic repeats [CRISPR]–CRISPR-associated protein 9 [CRISPR-Cas9]). They were found to have a cellular phenotype consistent with patient cells.⁵⁵ For FNIP1, previously generated and studied Fnip1 gene–targeted mice recapitulated the phenotype seen in patients.⁵⁶ The use of the ClinGen framework for novel gene–disease relationship in PAD highlights the relevance of comprehensive genetic analysis of patients and describes the utilization of functional evidence for proving disease causality and establishing gene–disease relationships. Methods such as genetic rescue, CRISPR-Cas9 gene editing, and nonhuman (eg, mouse or zebrafish) models have been invaluable in establishing the relevance of candidate gene defects to phenotype correction.^57,58

The ClinGen framework also allows for periodic reassessment and recuration of genes, thus enabling upgraded classifications that are based on new evidence, although IEIs/PIDs are rare to ultrarare, so a 3-year time frame may not always permit immediate reclassification. The ClinGen framework may also be evaluated and modified in the future for consideration of such ultrarare diseases and the burden of evidence required to establish valid gene–disease relationships, as this remains a dynamic process. The curation efforts of the AD-GCEP also encourage clinicians and scientists within the clinical immunology community to report in the literature candidate genes connected with PAD identified in patients to continuously improve our recognition and knowledge of these diseases while also providing additional data for curation and classification.

During the curation of these 74 PAD gene–disease relationships, it was observed that while there was heterogeneity in clinical presentation, there were also several consistent phenotypes across the genes. This is further supported by the fact that the 3 most common HPO terms used to define PAD are decreased circulating antibodies, pneumonia, and lymphadenopathy. Though these 3 terms are nonspecific in themselves, because they are identified in many gene–disease relationships, we were nonetheless able to create a dataset with a total of 3546 HPO terms. Further, for some genes (AICDA, NFKB1, NFKB2, PIK3CD, CARD11, BTK, SPI1, FNIP1), we identified specific HPO-term fingerprints. This curated database of terms can be potentially used in the future with artificial intelligence and machine learning tools, such as, among others, Phenomizer.⁵⁹

Finally, the harmonization of gene–disease nomenclature was another key contribution of the AD-GCEP. The AD-GCEP has generated a consensus nomenclature for the 74 gene–disease relationships that were evaluated in this curation process. The updated nomenclature would offer uniformity across multiple systems, such as OMIM, MONDO, and ClinGen. This nomenclature most accurately represents the gene–disease relationships used within the clinical immunology community. Ongoing collaboration between ClinGen GCEPs and other public databases are required to maintain the nomenclature harmonization effort.

This PAD curation effort presented herein represents the first tangible milestone of the Immunology CDWG. Other GCEPs and Variant Curation Expert Panels (VCEPs) within the Immunology CDWG are actively working to ultimately expand gene and variant curation for all the gene defects related to IEIs/PIDs. The current groups are represented by the Severe Combined Immunodeficiency GCEP/Combined Immunodeficiency GCEP and VCEP, the Primary Immune Regulatory Disorders GCEP, and the Hereditary Angioedema VCEP. Further, there is collaboration between the Immunology CDWG and other ClinGen GCEPs, such as the Kidney Complement GCEP, in addition to shared curation efforts with numerous other GCEPs. The other primary benefit and contribution of ClinGen is the rapid and immediate dissemination of curated data via the ClinGen database, enabling real-time access to clinicians and scientists worldwide.

The journey of curation and recuration of immunologically relevant genes and diseases will continue under the auspices of ClinGen, allowing therapeutic and diagnostic advances to be more rapidly implemented for patient care.

Key messages.

• Establishing the gene–disease validity of PAD is essential for appropriate diagnosis and management of these patients.

• The ClinGen gene curation framework offers a standardized approach to determining gene–disease relationships.

• Expansion of relevant terms related to PAD for future inclusion in the HPO database facilitates clinical diagnostic evaluation.

Abbreviations used

AD

Autosomal dominant

AD-GCEP

Antibody deficiencies GCEP

APDS

Activated p110δ syndrome

AR

Autosomal recessive

BENTA

B-cell expansion with NF-kB (nuclear factor kappa–light-chain enhancer of activated B cells) and T-cell anergy

BTK

Bruton tyrosine kinase

Cas9

CRISPR-associated protein 9

CDWG

Clinical Domain Working Group

ClinGen

Clinical Genome Resource

CRISPR

Clustered regularly interspaced short palindromic repeat

CVID

Common variable immunodeficiency

GCEP

Gene Curation Expert Panel

HPO

Human Phenotype Ontology (hpo.jax.org)

IEI

Inborn errors of immunity

IUIS

International Union of Immunological Societies

MONDO

Mondo Disease Ontology (mondo.monarchinitiative.org)

OMIM

Online Mendelian Inheritance in Man (www.omim.org)

PAD

Primary antibody deficiency

PCA

Principal component analysis

PID

Primary immunodeficiency

POU2AF1

POU domain class 2–associating factor 1

TRAF3

TNF receptor–associated factor 3

VCEP

Variant Curation Expert Panel

XL

X-linked

XLA

XL agammaglobulinemia