University of Ioannina
Multivariate analysis of genetic data on complex diseases and phenotypes
Abstract
dc:descriptionGenetic association studies mainly focus on associations between single genetic variants and single phenotypes. However, multivariate methods allow the investigation of genetic effects on multiple phenotypes and, therefore, can better capture the shared genetic architecture between correlated traits. These multi-trait approaches offer many advantages, such as increased statistical power for detecting novel genetic loci, reduced multiple testing burden and a more holistic exploration of genetic pleiotropy.The main objective of this thesis was to identify multivariate methods that utilise summary statistics from genome-wide association studies (GWAS), to compare and prioritise the most effective multivariate approaches for the discovery of novel genetic loci and exploration of pleiotropy. Specific aims included a systematic review of existing multivariate methods in order to identify the most efficient of them and practical applications of suitable multivariate methods on real GWAS data. This research applied advanced multivariate techniques to discover novel genetic loci for blood cell traits and prioritise genes with effects on different stages of haematopoiesis. It also examines the shared genetic architecture between inflammation and cardiometabolism and identified several pleiotropic loci and common mechanisms. Furthermore, the present thesis explored the genetic predisposition for comorbidity between Alzheimer's and cardiovascular diseases, identified novel loci and genes with pleiotropic effects and suggested new potential therapeutic targets.The systematic review on multivariate methods of single variant - multiple traits genome-wide associations using GWAS summary statistics identified 24 distinct methods of various statistical frameworks. Multi-Trait Association of GWAS (MTAG) was indicated as the most influential method within the scientific community and was prioritised due to its special feature to generate trait-specific summary statistics that can be used to post-GWAS analyses.Blood cells play a crucial role in the pathophysiology of several diseases, with their phenotypes exhibiting high heritability and polygenicity. However, previous GWAS have only partially explained the overall phenotypic variance of these traits. To further explore the genetic structure of blood cells, a multi-trait GWAS was conducted on fifteen blood cell traits, which identified 3,100 novel associations across 711 genomic loci, including 29 loci not previously linked to any blood trait in the European population. A subsequent gene expression colocalisation analysis highlighted variants with broad or cell type-specific effects. These findings provide insights into underlying mechanisms involved into different stages of haematopoiesis and consist potential therapeutic targets for blood-related disorders.C-reactive protein (CRP), a marker of chronic inflammation, is involved in a wide range of pathological conditions and likely shares biological pathways with metabolic traits. A multi-trait GWAS on CRP levels and cardiometabolic traits, including body mass inde, lipids and smoking, identified 41 novel CRP loci. A subsequent colocalisation analysis highlighted 41 shared variants between CRP and cardiometabolic risk factors, with 12 of them demonstrating unexpected discordant effects. Phenome-wide association studies linked these variants to clinical outcomes. These findings expand knowledge on the shared genetic architecture between inflammation and cardiometabolism, suggesting potential targets for preventive and therapeutic strategies.The co-occurrence of cardiovascular and neurodegenerative diseases, such as Alzheimer’s disease, indicates a shared genetic architecture between the two conditions. A multi-trait GWAS was performed on Alzheimer’s disease and several cardiovascular diseases, identifying 5 novel genetic loci for Alzheimer’s disease and 16 potentially pleiotropic loci affecting both conditions. Fine-mapping and colocalisation analyses suggested that lower expression of PLEC in the cardiac left ventricle or higher expression of C1Q genes can lead, through independent pathways, to increased risks of Alzheimer’s disease and atrial fibrillation. These findings suggest PLEC and C1Q as potential therapeutic targets for addressing the comorbidity of cardiovascular and neurodegenerative diseases.This thesis investigated theoretical foundations and practical applications of multivariate techniques in genetic epidemiology. It extends the current knowledge about the genetic architecture of several complex diseases and phenotypes, emphasising on the discovery of novel genetic loci and exploration of genetic pleiotropy.
Degree
thesis:*- Grantor dc:publisher
- University of Ioannina
- Year dc:date
- 2024
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Koskeridis, Fotios
- Κοσκερίδης, Φώτιος
Subjects
dc:subject × 24- Πολυμεταβλητή ανάλυση
- Ευρυγονιδιωματικές μελέτες συσχετίσεων
- Συστηματική ανασκόπηση
- Ευρυφαινομική μελέτη συσχετίσεων
- Γενετική ανάλυση συνεντοπισμού
- Μενδελιανή τυχαιοποίηση
- Multivariate analysis
- Genome-wide association studies (GWAS)
- Systematic review
- Phenome-wide association study (PheWAS)
- Colocalization analysis
- Mendelian randomization
- Βιολογία
- Φυσικές Επιστήμες
- Γενετική και Κληρονομικότητα
- Επιστήμες Υγείας
- Ιατρική και Επιστήμες Υγείας
- Επιδημιολογία
- Biological Sciences
- Natural Sciences
- Genetics and Heredity
- Health Sciences
- Medical and Health Sciences
- Epidemiology
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/59015