Recent technological advances and remarkable successes have led to genome-wide association studies (GWAS) becoming a tool of choice for investigating the genetic basis of common complex human diseases. These studies typically involve samples from thousands of individuals, scanning their DNA at up to a million loci along the genome to discover genetic variants that affect disease risk. Hundreds of such variants are now known for common diseases, nearly all discovered by GWAS over the last three years. As a result, many new studies are planned for the future or are already underway. In this thesis, I present analysis results from actual studies and some developments in theory and methodology. The Wellcome Trust Case Control Consortium (WTCCC) published one of the first large-scale GWAS in 2007. I describe my contribution to this study and present the results from some of my follow-up analyses. I also present results from a GWAS of a bipolar disorder sub-phenotype, and a recent and on-going fine mapping experiment. Building on methods developed as part of the WTCCC, I describe a Bayesian approach to GWAS analysis and compare it to widely used frequentist approaches. I do so both theoretically, by interpreting each approach from the perspective of the other, and empirically, by comparing their performance in the context of replicated GWAS findings. I discuss the implications of these comparisons on the interpretation and analysis of GWAS generally, highlighting the advantages of the Bayesian approach. Finally, I examine the effect of linkage disequilibrium on the detection and estimation of various types of genetic effects, particularly non-additive effects. I derive a theoretical result showing how the power to detect a departure from an additive model at a marker locus decays faster than the power to detect an association.

Genome-wide association studies (GWAS) exploit the correlation in ge- netic diversity along chromosomes in order to detect effects on disease risk without having to type causal loci directly. The inevitable downside of this approach is that, when the correlation between the marker and the causal variant is imperfect, the risk associated with carrying the predisposing allele is diluted and its effect is underestimated. This thesis explores four different facets of this risk dilution: (1) estimating true effect sizes from those observed in GWAS; (2) asking how the context of a GWAS, including the population studied, the genotyping chip employed, and the use of im- putation, affects risk estimates; (3) assessing how often the best-associated SNP in a GWAS coincides with the causal variant; and (4) quantifying how departures from the simplest disease risk model at a causal variant distort the observed disease risk model. Using simulations, where we have information about the true risk at the causal locus, we show that the correlation between the marker and the causal variant is the primary driver of effect size underestimation. The extent of the underestimation depends on a number of factors, including the population in which the study is conducted, the genotyping chip employed, whether imputation is used, and the strength, frequency, and disease model of the risk allele. Suppose that a GWAS study is conducted in a European population, with an Affymetrix 6.0 genotyping chip, without imputation, and that the causal loci have a modest effect on disease risk, are common in the population, and follow an additive disease risk model. In such a study, we show that the risk estimated from the most associated SNP is very close to the truth approximately two-thirds of the time (although we predict that fine mapping of GWAS loci will infrequently identify causal variants with considerably higher risk), and that the best-associated variant is very often perfectly or nearly-perfectly correlated with, and almost always within 0.1cM of, the causal variant. However, the strong correlations among nearby loci mean that the causal and best-associated variants coincide infrequently, less than one-fifth of the time, even if the causal variant is genotyped. We explore ways in which these results change quantitatively depending on the parameters of the GWAS study. Additionally, we demonstrate that we expect to identify substantial deviations from the additive disease risk model among loci where association is detected, even though power to detect departures from the model drops off very quickly as the correlation between the marker and causal loci decreases. Finally, we discuss the implications of our results for the design and interpretation of future GWAS studies.

La maladie de Crohn (MC) et la colite ulcéreuse (CU) sont des maladies inflammatoires chroniques du tube digestif qu’on regroupe sous le terme maladies inflammatoires de l’intestin (MII). Les mécanismes moléculaires menant au développement des MII ne sont pas entièrement connus, mais des études génétiques et fonctionnelles ont permis de mettre en évidence des interactions entre des prédispositions génétiques et des facteurs environnementaux - notamment la flore intestinale – qui contribuent au développement d’une dérégulation de la réponse immunitaire menant à l’inflammation de la muqueuse intestinale. Des études d’association pangénomiques et ciblées ont permis d’identifier plusieurs gènes de susceptibilité aux MII mais les estimations de la contribution de ces gènes à l’héritabilité suggèrent que plusieurs gènes restent à découvrir. Certains d’entre eux peuvent se trouver dans les régions identifiées par des études de liaison génétique. L’objectif de mon projet de doctorat était d’identifier un ou des facteurs de risque génétique dans la région chromosomale 19p (identifiée comme région de liaison IBD6) et de le/les caractériser au niveau fonctionnel. Nous avons d’abord entrepris une cartographie d’association de la région 19p. À la suite du génotypage successif de deux cohortes indépendantes, nous avons identifié un SNP intronique et quatre SNP codants dont un non-synonyme, rs8108738, tous localisés dans le gène microtubule associated serine threonine kinase gene-3 (MAST3) et associés aux MII. Peu d’information fonctionnelle sur MAST3 était disponible. Par contre MAST2, une protéine encodée par un gène de la même famille, régule l’activité du facteur de transcription inflammatoire NF-kappaB. Nous avons confirmé l’implication de MAST3 dans l’activité de NF-kappaB via un knockdown de MAST3 et des essais gène-rapporteur. Pour poursuivre la caractérisation fonctionnelle de MAST3, nous avons choisi une approche non ciblée pour étudier les effets de la variation des niveaux d’expression de MAST3 sur la cellule. C’est-à-dire que nous avons créé un 1er modèle cellulaire de surexpression du gène MAST3 dans les cellules HEK293 et analysé l’expression pangénomique endogène. La validation de l’expression génique dans un 2e modèle cellulaire de knockdown et de type cellulaire différent (THP1), nous a permis d’identifier et de contrer les effets non-spécifiques dus aux niveaux non-physiologiques. Notre étude d’expression a mené à l’identification d’un groupe de gènes dont l’expression est régulée par MAST3. Ces gènes sont majoritairement impliqués dans des fonctions immunitaires (cytokines pro-inflammatoires, régulateurs de NF-kappaB, migration cellulaire, etc.) et une forte proportion est régulée par NF-kappaB. Nous avons évalué l’importance du groupe de gènes régulés par MAST3 dans la présentation clinique des MII à travers des études d’expression dans des biopsies intestinales de patients atteints de CU. Nous avons constaté que l’expression de ces gènes est significativement supérieure dans les régions enflammées par rapport aux régions saines de la muqueuse intestinale des patients atteints de CU. Globalement, les résultats de nos études suggèrent que le facteur de risque aux MII MAST3 agit via la voie du facteur de transcription NF-kappaB pour influencer l’expression d’un groupe de gènes impliqués dans l’inflammation intestinale typique des MII. Chaque étude génétique sur les MII a le potentiel d’orienter les recherches fonctionnelles vers de nouvelles voies biologiques causales. Le dévoilement des mécanismes moléculaires sous-jacents à ces voies permet d’augmenter les connaissances sur le développement de ces maladies vers une compréhension plus complète de la pathogenèse qui permettra d’optimiser le diagnostic et le traitement de ces maladies. / Inflammatory bowel diseases (IBD) refer to different chronic inflammatory diseases of the digestive tract mainly Crohn’s disease (CD) and ulcerative colitis (UC). Mechanisms leading to the pathogenesis of IBD are not completely understood, but genetic and functional studies have highlighted interactions between genetic predispositions and environmental factors, such as the intestinal microbiota, as contributors to the deregulation of the immune response that leads to inflammation of the intestinal mucosa. Genome-wide and targeted association studies have identified several IBD susceptibility genes. However, estimations of the contribution of these genes to heritability of the disease suggest that more are to be discovered. Some of these genes may be in previously identified IBD linkage regions. The objective of my doctoral project was first, to identify risk factors in the 19p chromosomal region (IBD6), previously identified following a genome-wide linkage study, then to characterise them functionally. We first performed a comprehensive association mapping study of the 19p region. Our two-stage genotyping strategy led to the identification of one intronic SNP and four coding SNP –including one non-synonymous SNP, rs8108738 – all located in the microtubule associated serine threonine kinase gene 3 (MAST3) and associated to IBD. Very limited functional information on MAST3 was available at that time. However MAST2 (a gene in the same family as MAST3), is involved in the regulation of inflammation master switch, transcription factor NF-B. We confirmed the involvement of MAST3 in the modulation of NF-B via a knockdown of MAST3 and gene reporter assays. In order to further characterize the function of MAST3, we chose a non-targeted approach to study the effects of the modulation of MAST3 levels on the cell. More specifically, we created a 1st cell model of MAST3 overexpression in HEK293 cells and analysed the resulting genome-wide endogenous gene expression. Validation in a 2nd cell model consisting of a knockdown of MAST3 in THP1 cells, allowed to identify non-specific gene expression due to non-physiological MAST3 levels. Our expression study led to the identification of a group of genes whose expression is modulated by MAST3. These genes are mainly involved in immune functions (pro-inflammatory cytokines, NF-B regulation, cellular migration, etc) and a majority is regulated by transcription factor NF-B. We evaluated the importance of this MAST3-regulated gene set in the clinical manifestation of IBD through an expression study on biopsies of UC patients. We found that the expression the MAST3-regulated gene set was significantly enriched in inflamed region of the intestinal mucosa of UC patients compared to healthy region. Taken together, the results of our study suggest that IBD risk factor MAST3 acts on the NF-B pathway to influence the expression of a group of genes involved in intestinal inflammation typical of IBD. Every genetic study on IBD has the potential to lead functional research towards new causal biological pathways. The unravelling of the molecular mechanisms underlying these pathways aims to improve the comprehension of the pathogenesis of IBD and hopefully will allow for optimization of diagnostic and treatment of these diseases.

Résumé : La phase haploïde de la spermatogenèse (spermiogenèse) est caractérisée par une modification importante de la structure de la chromatine et un changement de la topologie de l’ADN du spermatide. Les mécanismes par lesquels ce changement se produit ainsi que les protéines impliquées ne sont pas encore complètement élucidés. Mes travaux ont permis d’établir la présence de cassures bicaténaires transitoires pendant ce remodelage par l’essai des comètes et l’électrophorèse en champ pulsé. En procédant à des immunofluorescences sur coupes de tissus et en utilisant un extrait nucléaire hautement actif, la présence de topoisomérases ainsi que de marqueurs de systèmes de réparation a été confirmée. Les protéines de réparation identifiées font partie de systèmes sujets à l’erreur, donc cette refonte structurale de la chromatine pourrait être génétiquement instable et expliquer le biais paternel observé pour les mutations de novo dans de récentes études impliquant des criblages à haut débit. Une technique permettant l’immunocapture spécifique des cassures bicaténaires a été développée et appliquée sur des spermatides murins représentant différentes étapes de différenciation. Les résultats de séquençage à haut débit ont montré que les cassures bicaténaires (hotspots) de la spermiogenèse se produisent en majorité dans l’ADN intergénique, notamment dans les séquences LINE1, l’ADN satellite et les répétions simples. Les hotspots contiennent aussi des motifs de liaisons des protéines des familles FOX et PRDM, dont les fonctions sont entre autres de lier et remodeler localement la chromatine condensée. Aussi, le motif de liaison de la protéine BRCA1 se trouve enrichi dans les hotspots de cassures bicaténaires. Celle-ci agit entre autres dans la réparation de l’ADN par jonction terminale non-homologue (NHEJ) et dans la réparation des adduits ADN-topoisomérase. De façon remarquable, le motif de reconnaissance de la protéine SPO11, impliquée dans la formation des cassures méiotiques, a été enrichi dans les hotspots, ce qui suggère que la machinerie méiotique serait aussi utilisée pendant la spermiogenèse pour la formation des cassures. Enfin, bien que les hotspots se localisent plutôt dans les séquences intergéniques, les gènes ciblés sont impliqués dans le développement du cerveau et des neurones. Ces résultats sont en accord avec l’origine majoritairement paternelle observée des mutations de novo associées aux troubles du spectre de l’autisme et de la schizophrénie et leur augmentation avec l’âge du père. Puisque les processus du remodelage de la chromatine des spermatides sont conservés dans l’évolution, ces résultats suggèrent que le remodelage de la chromatine de la spermiogenèse représente un mécanisme additionnel contribuant à la formation de mutations de novo, expliquant le biais paternel observé pour certains types de mutations. / Abstract : Germline mutations may arise from several endogenous and exogenous mechanisms in both male and female. However, recent next-generation sequencing (NGS) data confirmed that de novo mutations arise primarily in males. This observation suggests that specific spermatogenesis events are involved in the male mutation bias. One potential origin for male-driven mutations is the differentiation of spermatids into spermatozoa, which involves one of the most striking and global chromatin remodeling processes, where histone-bound chromatin is converted into highly condensed protaminated DNA toroid. Using pulse-field gel electrophoresis and comet assay on flow cytometry sorted cells, it was established that chromatin remodeling process is characterized by a transient surge in DNA double strand breaks (DSBs) in the whole population of murine spermatids, which get repaired by the end of spermiogenesis. Using a highly active nuclear extract and immunofluorescences, topoisomerases and markers of DNA repair systems were shown at these steps. Since haploid cells cannot rely on homologous recombination for templated DNA repair, it was hypothesized that this process may be genetically unstable and largely responsible for the observed male de novo mutations bias. Although very challenging, a method allowing the specific genome-wide mapping of DSBs using NGS was developed to establish the genomic distribution of DSBs during chromatin remodeling. It was shown that intergenic regions were enriched in DSBs, particularly LINE1, satellite DNA and simple repeats. Motif finding on potential hotspots showed that proteins from FOX and PRDM families may be implicated. Although homologous recombination cannot take place during spermiogenesis, an enrichment in BRCA1 motif was found, which is also known to be implicated in NHEJ and removal of topoisomerase adducts. Topoisomerase-like SPO11 motif was also enriched suggesting that the meiotic machinery may also be implicated during chromatin remodeling. Moreover, although DSBs tend to accumulate in intergenic regions, gene ontology analysis of hotspot-containing genes showed a marked enrichment in genes related to neurons and brain development. This result hence supports the fact that neurological disease associated mutations are also male biased and associated with advanced paternal age. Since DSB formation during spermiogenesis is conserved through evolution, these results suggest that chromatin remodeling in spermatids represents a significant component in the reported male de novo mutation bias.

<p>Our primary goal is to better understand complex diseases using statistically disciplined approaches. As multi-modal data is streaming out of consortium projects like Genotype-Tissue Expression (GTEx) project, which aims at collecting samples from various tissue sites in order to understand tissue-specific gene regulation, new approaches are needed that can efficiently model groups of data with minimal loss of power. For example, GTEx project delivers RNA-Seq, Microarray gene expression and genotype data (SNP Arrays) from a vast number of tissues in a given individual subject. In order to analyze this type of multi-level (hierarchical) multi-modal data, we proposed a series of efficient-score based tests or score tests and leveraged groups of tissues or gene isoforms in order map genomic biomarkers. We model group-specific variability as a random effect within a mixed effects model framework. In one instance, we proposed a score-test based approach to map expression quantitative trait loci (eQTL) across multiple-tissues. In order to do that we jointly model all the tissues and make use of all the information available to maximize the power of eQTL mapping and investigate an overall shift in the gene expression combined with tissue-specific effects due to genetic variants. In the second instance, we showed the flexibility of our model framework by expanding it to include tissue-specific epigenetic data (DNA methylation) and map eQTL by leveraging both tissues and methylation. Finally, we also showed that our methods are applicable on different data type such as whole transcriptome expression data, which is designed to analyze genomic events such alternative gene splicing. In order to accomplish this, we proposed two different models that exploit gene expression data of all available gene-isoforms within a gene to map biomarkers of interest (either genes or gene-sets) in paired early-stage breast tumor samples before and after treatment with external beam radiation. Our efficient score-based approaches have very distinct advantages. They have a computational edge over existing methods because they do not need parameter estimation under the alternative hypothesis. As a result, model parameters only have to be estimated once per genome, significantly decreasing computation time. Also, the efficient score is the locally most powerful test and is guaranteed a theoretical optimality over all other approaches in a neighborhood of the null hypothesis. This theoretical performance is born out in extensive simulation studies which show that our approaches consistently outperform existing methods both in statistical power and computational speed. We applied our methods to publicly available datasets. It is important to note that all of our methods also accommodate the analysis of next-generation sequencing data.</p> / Dissertation

The population level bias towards right-handedness in humans implies left-hemisphere dominance for fine motor control. Left-handedness and reduced cerebral asymmetry have been linked to neurodevelopmental disorders such as dyslexia. Understanding the biology of these traits at a genetic level is crucial for understanding the relationship between handedness and neurodevelopmental disorders. Here I present genome-wide association study (GWAS) meta-analyses for both relative hand skill (handedness, n = 728) and reading-related traits (n = 548) in individuals with dyslexia. I uncovered a genome-wide significant association in an intron of PCSK6 associated with relative hand skill. PCSK6 is a protease that cleaves NODAL proprotein into an active form, and NODAL determines the development of left/right (LR) asymmetry in bilaterians. I performed pathway analyses of the GWAS data that revealed handedness is determined in part by the mechanisms that establish left/right (LR) asymmetry early in development, such as NODAL signalling and ciliogenesis. This finding replicated in a general population cohort unaffected with neurodevelopmental disorders (n = 2,666). A key stage in LR asymmetry development is the rotation of cilia that creates a leftward flow of NODAL. Candidate genes for dyslexia are involved in both neuronal migration and ciliogenesis. Ciliopathies can cause not only LR body asymmetry phenotypes, but also cerebral midline phenotypes such as an absent corpus callosum. Furthermore, I identified a genome-wide significant association with non-word reading located in an intron of MAP1B, a gene involved in neuronal migration that causes an absent corpus callosum when disrupted in mice. However, this finding did not replicate in two independent cohorts with dyslexia (n = 156 & 199), or in the general population cohort (n = 2,359). Though these cohorts had inadequate reading measures and poorly matched ascertainment for dyslexia. I also performed copy number variation (CNV) pathway and burden analyses of 920 individuals with dyslexia and 1,366 unselected controls, but did not find that rare CNVs play a major role in the etiology of dyslexia. Based on these results I propose that common variants in genes responsible for ciliogenesis and corpus callosum development influence traits such as handedness and reading ability.

The post human genome sequence era has begun to explore various aspects of the functional genome in relation to disease including gene expression, genetic variation and epigenetics. The genetic determinants of common and complex phenotypes are difficult to resolve even though their heritability is recognised. Recent genome-wide association studies (GWAS) for common diseases has identified many new disease susceptibility associated loci. These loci often lie in non-coding regions of the genome and disease associated genetic variants are proposed to act by modulating gene expression. This thesis investigated genetic variation as determinants of gene expression in the context of the immune system especially focused on the innate immune and inflammatory responses. Different primary human immune cell types were collected from healthy volunteers of European ancestry to achieve this. In order to identify genetic variants associating with gene expression, expression quantitative trait loci (eQTL) mapping was conducted in a cell type specific manner. The primary dataset (n=288) consists of CD19⁺ B-cells from the adaptive immune system and CD14⁺ monocytes from the innate immune system. 78% of the total cis eQTL were found to be cell type specific and include genes relating to their roles in the immune response. Trans eQTL showed greater cell type specificity and include master regulatory eQTL on the LYZ locus at chromosome 12q15 in monocytes and the KLF4 (9p31) in B-cells. The identified eQTL are implicated in association with autoimmune disease susceptibility including inflammatory bowel disease, diabetes and rheumatoid arthritis. The second analysed dataset (n=64) consists of CD14+ monocytes and macrophages differentiated ex vivo. Macrophages are involved in many inflammatory diseases as well as in the innate immune response. The differential gene expression and eQTL mapping analyses were conducted to investigate macrophages specific gene expression signatures and associations to genetic variants. Macrophage eQTL are involved in signal transduction for the inflammatory response (IL1RN and STAT4) and lipid metabolism (PPARG) with implication for metabolic disease association. The eQTL analyses using primary immune cell types provide insights into genetic variation in association to gene expression which is involved in autoimmunity and disease susceptibility.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterised clinically by rapidly progressive paralysis leading ultimately to death from respiratory failure. There is no cure for ALS and no definitive explanation for the onset and rapid progression of motor neuron degeneration. Genetics is a known risk factor for a portion of familial cases. However, the role of genetics in the commoner sporadic form of the disease is poorly understood, although numerous genes have been implicated. The primary aim of this thesis project is to uncover the genetic causes that underlie ALS. To accomplish this goal, the main focus of this thesis is to perform genome-wide association analysis of sporadic ALS using high density SNP arrays. This thesis describes the first and the largest genome-wide association studies of ALS to date. Results demonstrate that there is no single large effect susceptibility variant underlying a large proportion of ALS, such as ApoE in Alzheimer’s disease. However, the genotyping data has been made publically available and the digital nature of this data means that it is a resource that can grow with future studies. Beyond genome-wide association, this thesis describes work using linkage, haplotype and sequence analysis to investigate the genetic overlap between ALS and frontotemporal dementia. Lastly, this thesis presents a novel method for linkage analysis using high throughput SNP arrays. Ultimately, it is hoped that by uncovering the genes that cause ALS, such knowledge will shed light on the pathogenic mechanisms underlying motor neuron degeneration and potentially lead to new rational therapies effective in slowing or even halting disease progression.

The aim of this thesis is to obtain a better understanding of mutation rates within as well as between the genomes of humans and chimpanzees using data generated by high throughput sequencers. I will start with a review of the field and an overview of the technologies and protocols used to generate and analyse high throughput sequencing data. I apply some of the discussed techniques to show that there is evidence of a selective advantage of pathogenic de novo mutations in the Fibroblast Growth Factor Receptor 3 gene in the male germ line of humans. Furthermore, I use some of the methods to generate a map of genome-wide sequence variation in Western chimpanzees. Ever since Darwin [Darwin, 1871] and Huxley [Huxley, 1863] postulated more than a century ago that African great apes are our closest living evolutionary relatives, the study of chimpanzee individuals is of great scientific interest from an evolutionary point of view, as comparisons between the genomes of human and chimpanzee offer the potential to help to understand the molecular basis for similarities and differences between the two species. I use the generated data to explore the breadth of the nucleotide diversity in the chimpanzee genome in order to shed light on whether or not the local variation in mutation rate has been conserved since the divergence of the two species and to place human nucleotide diversity into perspective with an evolutionary closely related species. I explore the relationship of nucleotide diversity in chimpanzees with specific large-scale genome features to reveal a number of highly significant correlations which explain over 40% of the observed variation. I use data from the 1000 Genomes Project to examine the occurrence of ancestral polymorphisms shared between human and chimpanzee on a genome-wide scale. These ancestral polymorphisms do not only influence fine-scale divergence rates across the genome in very closely related species, they are also good candidates for regions under balancing selection and thus, they are a useful tool to study long-time population demographics and speciation. Using these variants, I postulate that long-term balancing selection may be more common than previously believed. I conclude with a discussion of the results contained in the body of the thesis and suggest a number of areas for future research.

The fundamental theme my thesis explores is the relationship between genetic variation and phenotypic variation. It addresses three main questions. What is the genetic architecture of traits in the HS? How can sequence information help identifying the sequence variants and genes responsible for phenotypic variation? Are the genetic factors contributing to phenotypic variation in the rat homologous to those contributing to variation in the same phenotype in the mouse? To address these questions, I analysed data collected by the EURATRANS consortium on 1,407 Heterogeneous Stock (HS) rats descended from eight inbred strains through sixty generations of outbreeding. The HS rats were genotyped at 803,485 SNPs and 160 measures relevant to a number of models of disease (e.g. anxiety, type 2 diabetes, multiple sclerosis) were collected. The eight founders of the Stock were genotyped and sequenced. I identified loci in the genome that contribute to phenotypic variation (Quantitative Trait Loci, QTLs), and integrated sequence information with the mapping results to identify the genetic variants underlying the QTLs. I made some important observations about the nature of genetic architecture in rats, and how this compares to mice and humans. I also showed how sequence information can be used to improve mapping resolution, and in some cases to identify causal variants. However, I report an unexpected observation: at the majority of QTLs, the genetic effect cannot be accounted for by a single variant. This finding suggests that genetic variation cannot be reduced to sequence variation. This complexity will need to be taken into account by studies that aim at unravelling the genetic basis of complex traits.