Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders
Genome-wide association studies (GWAS) have identified numerous loci associated with diverse human traits. While these discoveries offer valuable insights into disease biology and inform patient care, key challenges remain. These include the disconnect between identified variants and biological mech...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis |
| Language: | English English |
| Published: |
Department of Pathology
2025
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| _version_ | 1867614272760053760 |
|---|---|
| access_status_str | Open Access |
| author | Defo, Joel |
| author2 | Ramesar, Rajkumar |
| author_browse | Defo, Joel Ramesar, Rajkumar |
| author_facet | Ramesar, Rajkumar Defo, Joel |
| author_sort | Defo, Joel |
| collection | Thesis |
| description | Genome-wide association studies (GWAS) have identified numerous loci associated with diverse human traits. While these discoveries offer valuable insights into disease biology and inform patient care, key challenges remain. These include the disconnect between identified variants and biological mechanisms, the difficulty in detecting low-frequency or moderate-effect variants, and the inherent complexity of most human phenotypes. To address these limitations and enhance discovery power, large sample sizes are increasingly employed through individual study expansion or meta-analysis. However, the classical single-variant approach, analyzing individual genes or variants in isolation, may not fully capture the intricate genetic architecture of complex diseases. This can be due to genetic heterogeneity or limitations associated with gene-based analysis, where power is lost due to non-effect variants within a gene or low-frequency causal variants. Consequently, traditional GWAS often lack robust mechanistic insights into the functional underpinnings of complex traits. This thesis proposes a novel framework that integrates GWAS signals across multiple studies, leveraging summary statistics to improve association mapping and detect weak signals missed in single GWAS. The method aggregates SNP-level signals to the gene level and incorporates protein-protein interaction (PPI) networks for association mapping at gene and pathway/subnetwork levels. We applied this framework to GWAS summary statistics from seven European bipolar disorder cohorts. The analysis identified seven genes, including a highly significant effect at the gene level for AGT. Additionally, a significant subnetwork was identified with Estrogen Receptor 1 (ESR1) as the central hub. Furthermore, the method was employed to explore the genetic overlap between suicide and psychiatric disorders in the FinnGen database. This analysis revealed disease-specific traits sharing common risk factors with suicide. Notably, several genes with small effects were identified at both the gene and subnetwork levels, highlighting the involvement of overlapping genes, pathways, and subnetworks in the underlying molecular mechanisms. Functional enrichment analysis of hub genes based on annotations from Reactome and KEGG databases revealed significant pathways, with top significant involved in signaling, nervous system development, and the immune system. Finally, the framework was applied to investigate the genetic overlap between suicidality and subcortical brain volume. This analysis identified potentially significant genes, hub genes with small effects, along with a network of interacting genes. Enrichment analysis of the network genes revealed pathways with top significance associated with signaling, immune function, and nervous system development. In conclusion, the method presented in this thesis provides new insights into functional and molecular mechanisms issued from a gene/subnetwork-centric approach and exhibits candidate genes for drug targets and drug re-purposing. Altogether, these findings point to important pathways and possible regulatory mechanisms that may be involved in the emergence of disorders, as well as the multi- and co-morbidities associated with them. Clinical research and an integrated strategy of additional modalities will advance individualized mechanistic understandings of these complex illnesses. |
| format | Thesis |
| id | oai:open.uct.ac.za:11427/42196 |
| institution | University of Cape Town (South Africa) |
| language | English eng |
| last_indexed | 2026-06-10T12:49:24.885Z |
| license_str | Not specified — see source repository |
| provenance_str_mv | Harvested via OAI-PMH from UCTD — University of Cape Town Open Access Repository |
| publishDate | 2025 |
| publishDateRange | 2025 |
| publishDateSort | 2025 |
| publisher | Department of Pathology |
| publisherStr | Department of Pathology |
| record_format | dspace |
| source_str | UCTD — University of Cape Town Open Access Repository |
| spelling | oai:open.uct.ac.za:11427/42196 Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders Defo, Joel Ramesar, Rajkumar Genetic disorders Genome-wide association studies (GWAS) have identified numerous loci associated with diverse human traits. While these discoveries offer valuable insights into disease biology and inform patient care, key challenges remain. These include the disconnect between identified variants and biological mechanisms, the difficulty in detecting low-frequency or moderate-effect variants, and the inherent complexity of most human phenotypes. To address these limitations and enhance discovery power, large sample sizes are increasingly employed through individual study expansion or meta-analysis. However, the classical single-variant approach, analyzing individual genes or variants in isolation, may not fully capture the intricate genetic architecture of complex diseases. This can be due to genetic heterogeneity or limitations associated with gene-based analysis, where power is lost due to non-effect variants within a gene or low-frequency causal variants. Consequently, traditional GWAS often lack robust mechanistic insights into the functional underpinnings of complex traits. This thesis proposes a novel framework that integrates GWAS signals across multiple studies, leveraging summary statistics to improve association mapping and detect weak signals missed in single GWAS. The method aggregates SNP-level signals to the gene level and incorporates protein-protein interaction (PPI) networks for association mapping at gene and pathway/subnetwork levels. We applied this framework to GWAS summary statistics from seven European bipolar disorder cohorts. The analysis identified seven genes, including a highly significant effect at the gene level for AGT. Additionally, a significant subnetwork was identified with Estrogen Receptor 1 (ESR1) as the central hub. Furthermore, the method was employed to explore the genetic overlap between suicide and psychiatric disorders in the FinnGen database. This analysis revealed disease-specific traits sharing common risk factors with suicide. Notably, several genes with small effects were identified at both the gene and subnetwork levels, highlighting the involvement of overlapping genes, pathways, and subnetworks in the underlying molecular mechanisms. Functional enrichment analysis of hub genes based on annotations from Reactome and KEGG databases revealed significant pathways, with top significant involved in signaling, nervous system development, and the immune system. Finally, the framework was applied to investigate the genetic overlap between suicidality and subcortical brain volume. This analysis identified potentially significant genes, hub genes with small effects, along with a network of interacting genes. Enrichment analysis of the network genes revealed pathways with top significance associated with signaling, immune function, and nervous system development. In conclusion, the method presented in this thesis provides new insights into functional and molecular mechanisms issued from a gene/subnetwork-centric approach and exhibits candidate genes for drug targets and drug re-purposing. Altogether, these findings point to important pathways and possible regulatory mechanisms that may be involved in the emergence of disorders, as well as the multi- and co-morbidities associated with them. Clinical research and an integrated strategy of additional modalities will advance individualized mechanistic understandings of these complex illnesses. 2025-11-12T11:40:18Z 2025-11-12T11:40:18Z 2025 2025-11-12T10:45:04Z Thesis / Dissertation Doctoral PhD http://hdl.handle.net/11427/42196 en eng application/pdf Department of Pathology Faculty of Health Sciences University of Cape Town |
| spellingShingle | Genetic disorders Defo, Joel Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| thesis_degree_str | Doctoral |
| title | Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| title_full | Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| title_fullStr | Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| title_full_unstemmed | Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| title_short | Leveraging gene/subnetwork meta-analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| title_sort | leveraging gene subnetwork meta analysis to recover signal and deconvolute the interactions between genes in the risk of genetic disorders |
| topic | Genetic disorders |
| url | http://hdl.handle.net/11427/42196 |
| work_keys_str_mv | AT defojoel leveraginggenesubnetworkmetaanalysistorecoversignalanddeconvolutetheinteractionsbetweengenesintheriskofgeneticdisorders |