Causal Gene Prioritization Across Diverse Diseases Through Multi-Omic Data Integration

Genome-wide association studies (GWASs) have identified thousands of disease susceptibility loci, yet the underlying causal variants, genes and tissues of action are unknown for most of the reported associations. This limits our biological understanding of the mechanisms underlying diseases and presents a major bottleneck for experimental follow up and clinical translation of the findings. Technological advances now enable in-depth, high throughput multi-omic assessment (e.g. genomic sequencing, plasma proteomics) in large-scale patient and population studies. This thesis explores the application and integration of multi-omic data for the prioritization of effector genes underlying disease development, including (i) analysis of whole exome-sequencing (WES) data to test the role of rare coding variants using fat distribution and its effect on cardiometabolic risk as a use case, (ii) integration of proteogenomic with phenome-wide data to systematically prioritise candidate causal genes across diverse human diseases and (iii) testing the importance of sex differences in these studies and their findings. Firstly, I conducted exome-wide gene burden analyses of rare, loss of function (LoF) variants in 184,246 participants of UK Biobank study and identified five genes (*PLIN1*, *INSR*, *ACVR1C PDE3B*, *PLIN4*) associated with body fat distribution, assessed by waist-to-hip ratio adjusted for body mass index (WHRadjBMI). LoF variants of *PLIN4*, *INSR* and *PDE3B* showed significantly larger standardized effect sizes for females compared to males in sex-stratified gene-based analyses. The phenotypic follow up analysis I have performed on the LoF variants of the identified genes highlighted the importance of understanding the mechanisms underlying fat distribution beyond total body fat for cardiometabolic disease risk. Secondly, I have used measurements from a novel antibody-based proteomics assay to study the variation in and nearby protein-encoding genes (described as being in cis) driving translation into differential plasma levels of the gene product for over 3,000 protein targets, some of which had not been previously targeted, contributing a valuable resource to the community. As an alternative approach in effector gene prioritization, I leveraged this cis-based proteogenomic knowledge to identify disease causing genes and proteins across the human phenome, providing human genetic support for therapeutic targets (e.g. as gastrin-releasing peptide for type 2 diabetes) and improving causal gene assignment at 40% (n=192) of overlapping GWAS loci. Building upon my previous work, I assembled and meta-analysed proteogenomic data from 38 international studies to conduct the largest antibody-based proteogenomic study (maximum sample size =78,664) to date, delivering an extensive pQTL catalogue for use by the scientific community, assessing sources of heterogeneity in proteogenomic studies, demonstrating limited convergence of cis- and trans-pQTL based causal inference methods and providing examples with strong evidence of causal genes for diverse diseases. Finally, I analysed differences in proteogenomic associations between 30,307 females and 26,058 males for 5,100 proteins across two technologies and demonstrated strongly conserved genetic effects for the vast majority of proteins between sexes, with few biologically plausible exceptions, in spite of large differences in plasma levels. In summary, this thesis provides concrete examples of how integration of omics data can facilitate prioritization of causal genes for diverse range of complex diseases across clinical specialities to improve our understanding of biological mechanisms underlying these diseases and facilitate rational target prioritization or drug repurposing opportunities.

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

• Koprulu, Mine

Advisors dc:contributor.advisor

• Langenberg, Claudia
• Wareham, Nicholas

• causal gene prioritization
• complex diseases
• genetic epidemiology
• genomics
• multi-omics
• proteomics
• rare variants

Licence

• https://www.repository.cam.ac.uk/bitstreams/e67a3851-a8af-4929-b3e6-442300bc6739/download
• https://www.rioxx.net/licenses/all-rights-reserved/

Language dc:language

eng

DOI dc:identifier.doi

https://doi.org/10.17863/CAM.112533

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/374472