Common variants in antibody deficiencies

The inborn errors of immunity (IEIs) comprise a group of almost 500 diseases charac- terised by immune dysfunction of genetic origin. Many of these disorders represent canoni- cal examples of Mendelian disease and study of their genetic aetiology is largely conducted within a rare-variant, monogenic paradigm. However, most cases of the two most common IEIs, selective IgA deficiency (SIgAD) and common variable immunodeficiency (CVID), have no known Mendelian origin. In this work I investigate the common-variant architec- ture of SIgAD and CVID through genome-wide association studies (GWAS) and use of the pleiotropy-informed conditional false discovery rate (cFDR), which allows one to leverage information from highly-powered studies of related traits. Application of the cFDR requires identification of ‘auxiliary’ traits which are genetically related with the primary trait. Estimation of the genetic correlation is unsuitable for this task in the small-sample regime due to a lack of power. I improve an existing nonparametric test of genetic similarity, the genome-wide pairwise-association signal sharing (GPS) test, through identification of a more suitable null distribution to achieve uniformity of p-values under the null. I devise a permutation testing procedure to estimate the parameters of this distribution and implement a performant algorithm for fast bivariate ecdf evaluation. I demonstrate an advantage for the GPS test in the small-sample setting over tests of non- zero genetic correlation and the nonparametric comparator Hoeffding’s test using simulated case-control GWAS and a panel of disease traits from the UK Biobank. To conventionally maximise the power of GWAS-based discovery, I conduct the largest meta-analysis to date of SIgAD and CVID, compiling publicly available summary statistics with novel GWAS using data from the UK Biobank and the ‘Intrepid’ Study’s IEI cohort. I report four new associations for SIgAD and one for CVID, buttressing the case for a poly- genic, common-variant aetiology for both diseases. I supplement this with meta-analyses of serum IgA, IgG, and IgM as ostensibly relevant physiological traits; the serum IgA GWAS I present is the largest to date and features eight novel associations. I use the genetic correlation and the GPS test to measure the coheritability of CVID and SIgAD with other immune-mediated diseases and traits, identifying serum IgA, asthma, and rheumatoid arthritis as suitable auxiliary traits for use in the cFDR procedure. I use the procedure to identify 18 additional SIgAD and 10 additional CVID associations, noting considerable overlap among candidate genes for these associations with the genes known to harbour IEI-related rare variants. Finally I explore the shared genetic architecture of CVID, SIgAD, and serum antibody production production, and I adduce my findings in an argument for the relevance of common-variant analysis in the study of IEIs.

Name dc:type.qualificationname

Doctor of Philosophy (PhD)

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

University of Cambridge

Year dc:date.issued

2024

Author dc:creator

• Willis, Thomas

Advisor dc:contributor.advisor

• Wallace, Chris

• biostatistics
• GWAS
• pleiotropy
• statistical genetics

Licence

• https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/7d8146b3-f131-45a6-878c-01650a985eac/download
• https://creativecommons.org/licenses/by/4.0/

Language dc:language

eng

Author Identifier

0000-0001-6138-3496

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/379280