Dissertations / Theses on the topic 'CNV'

This master thesis deals with analysis of structural variation of genome and with methods of its sequencing across all generations. Subsequently it contains a description of copy number variation and methods of its detection. The experimental part focuses on algorithm proposal for CNV detection according analysis and testing of uneven coverage in genome, variable representation of GC content and distance of sequence reads. Finally, the algorithm for detecting copy number variation is tested on genomic data of bacteria Klebsiella pneumoniae.

Copy number variation detection in prokaryotic organisms is currently receiving more and more attention, mainly due to the association of CNV with pathogenicity and antibiotic resistance in bacteria. The algorithm designed in this thesis uses peak detection in sequencing coverage to detect CNV segments. Read coverage is commonly obtained by mapping sequencing reads of one individual to an already known reference of another individual of the same species. However, two individuals will always differ in a certain number of genes, resulting in unmapped reads that are unnecessarily discarded. Therefore, this work assumes that the biological accuracy of CNV detection can be increased by using a new reference that is created from the same set of reads as the reads mapped to this reference. Sequencing reads of Klebsiella pneumoniae individuals are used to verify this assertion.

Humans have highly variable number of alpha defensin genes, with between 3-16 diploid copies. Alpha-defensin genes have important roles in human innate immunity and diseases. Recently, GWAS studies reported this locus associated with IgA nephropathy and periodontitis. However, the underlying mechanism of association is not clear. In this Ph.D. thesis, human alpha defensin CNV flanking haplotype diversity in global populations was studied and the association between diseases and haplotype classes was discussed. Then a novo method to detect variants from inside the DEFA1A3 CNV was developed and a list of potential disease-related mutations for further functional studies was generated. The association between CNV internal variants and flanking haplotype classes was studied. Non-allelic homologous recombination was found to be the major mechanism of CNV formation of alpha defensin CNV. Analysis results were verified by PCR and Sanger sequencing-based methods. Additional to that, the haplotype diversity analysis highlighted an unusual haplotype 5T/7C which is only found in European populations but highly diverged from other human haplotypes. Further evidence was provided to suggest that this is an introgressed haplotype from Neanderthals. Furthermore, we used Oxford Nanopore to reconstruct haplotype structure in DEFA1A3 CNV and discussed its advantages and limitations by our analysis results. In brief, this Ph.D. research greatly improved our understanding of DEFA1A3 global diversity, evolutionary history, diseases and haplotype association.

El experimento clásico evocador de la onda CNV consiste en la presentación consecutiva de dos estímulos. El primero advierte al sujeto de la aparición de un segundo estímulo, tras el que deberá realizar algún tipo de acción, generalmente motora (Walter et als., 1964). La onda eléctrica registrada en el intervalo de estos dos estímulos, mediante la utilización de electrodos situados en la superficie del cuero cabelludo, es la que se identifica como macrocomponente CNV.

Todos los registros de fenómenos bioeléctricos realizados en el ser humano se enfrentan con un problema fundamental: la presencia de ruido acompañando a la señal eléctrica que se desea registrar. En el caso de los potenciales relacionados con el evento este problema resulta ser extremadamente complicado, debido a que la razón señal/ruido (SNR), o sea el cociente de las variabilidades presentadas por ambas, suele ser extremadamente bajo.

Tradicionalmente el aumento de la SNR se ha solventado registrando gran cantidad de ensayos individuales, y promediando los valores obtenidos en cada punto "t" de la serie temporal. Este procedimiento, comúnmente utilizado en la mayoría de laboratorios, presenta graves inconvenientes de aplicación por lo que respecta al complejo CNV. En primer lugar cabe destacar la importante contaminación que sufre este multicomponente como resultado de los movimientos oculares producidos por el sujeto (Hillyard y Galambos, 1970). Esto hace que, si el rechazo de los ensayos contaminados por artefactos se produce simultáneamente a la obtención del registro, pueda dilatarse excesivamente la sesión experimental, con objeto de conseguir un número de ensayos suficientes en base a los que poder obtener la estimación de la señal. La producción de un número excesivo de ensayos comporta, a su vez, fatiga en el sujeto, resultando de esta forma bastante difícil el registro de la señal en población psicopatológica o en niños y ancianos. Por otra parte, si la selección de ensayos libres de movimientos oculares u otros artefactos se realiza con posterioridad a la finalización de la sesión de registro, es posible que el número de ensayos finalmente seleccionados sea tan reducido que impida la correcta estimación de la señal mediante el promediado simple, puesto que no serán suficientes para magnificar de forma conveniente la SNR.

Siguiendo la sugerencia realizada por Tukey (1978), se propone en esta tesis la utilización, conjuntamente con el promediado, de alguna técnica de suavizado (Tukey, 1977; Velleman y Hoaglin, 1981). La aplicación de estas técnicas produce un cambio en la perspectiva de la estimación de la señal, sustituyendo el promediado transversal clásico que utiliza el promediado simple por otro longitudinal.

Las técnicas de suavizado utilizadas son las contempladas por el Análisis Exploratorio de Datos (EDA) y comparten, por tanto, sus propiedades y ventajas, que pueden resumirse en:

- sencillez de formalización
- utilización de índices resistentes y robustos
- análisis gráfico
- análisis de los residuales

La técnica se plantea conseguir una descripción simple de variable "v" (diferencia de potencial) en función del tiempo, descomponiéndose cada valor (Dato = Parte Suavizada + Parte Rugosa). La parte suavizada no pretende ser una descripción mediante una fórmula sino simplemente una curva alisada que recoja, a gran escala, la estructura de la secuencia de datos, y por consiguiente la parte rugosa sea un proceso aleatorio.

Este tipo de alisado aplicado a una secuencia de datos EEG actúa como un filtro de pasa-bajos, por lo tanto elimina los componentes que presentan frecuencias altas, siendo a nuestro entender este hecho lo que los hace especialmente adecuados para el análisis de componentes lentos como la CNV. Ciertamente este proceso dependerá básicamente de dos parámetros, la amplitud de ventana empleada y la ponderación de los elementos que la componen.

La utilización de las técnicas de suavizado, al actuar sobre las respuestas eléctricas de alta frecuencia, permite eliminar aquellos ensayos contaminados que no hayan sido mitigados mediante el uso del filtrado analógico, o cualquier otro método de rechazo de artefactos.

La aplicación de los alisadores se realizará de forma exploratoria, o sea, observando la actuación de diferentes tipos de éstos sobre la serie registrada. Esto no puede ser de otra forma debido a que no es conocida la función de transferencia que permitiría el cambio del análisis en el dominio del tiempo al dominio de la frecuencia, siendo ésta, en la actualidad, su principal limitación.

Se ha realizado una aplicación de las técnicas de suavizado a unos datos obtenidos en experimentos CNV y RP (Potencial de Preparación). Tanto en los experimentos CNV como los RP la aplicación combinada de alisadores y el promediado simple proporciona una estimación de la señal a partir de un reducido número de ensayos individuales, evitando de esta forma la aparición de los procesos de fatiga y habituación. Esta es una de las grandes ventajas en comparación con el promediado simple, que requiere un mayor número de ensayos para obtener una estimación similar.

Esta mejora en la estimación redundará en la de los posibles análisis realizados utilizando este componente corno variable dependiente, en estudios de relación con otras variables experimentales con las que puede encontrarse relacionado.

La utilización de alisadores demasiado fuertes, corno por ejemplo la regresión Lowess, afecta a componentes corno el P300 y el RP deformando excesivamente su estimación.

En el apéndice se realiza un estudio de simulación en el que parece confirmarse la mejora que supone la realización de un proceso de suavizado de forma complementaria al del promediado. En un primer estadio se constató que en caso de que las SNR se aproxime a la unidad, el suavizado consigue una correcta estimación utilizando únicamente un sólo ensayo, incluso con los alisadores considerados más blandos. Ciertamente, en el caso de los registros ERP no es frecuente encontrar SNR's tan elevadas, ni siquiera en el caso de la CNV, que quizás sea uno de los componentes en el cual ésta es de las más altas. Por este motivo se trabajó con una SNR más próxima a la que suele ser habitual en este tipo de macrocomponente. En esta situación se constató que, si bien la estimación en base al ensayo individual resultó muy deformada, sí que es posible la correcta detección de la señal en base a un menor número de ensayos, tanto si se suaviza el promedio de éstos, como si se calcula el promedio de los mismos ensayos previamente suavizados.
The classic CNV evoking experiment consists on the consecutive presentation of two stimuli. First advises the subject the forthcoming second stimulus, after which some kind of action has to be done, often motor (Walter et als., 1964). The registered electric wave in the interval between these two stimuli, by using scalp electrodes, is named as CNV macrocomponent.

Every bioelectric phenomena registered on the human being are confronted to a main problem: the presence of noise going along with the electric signal intended to register. Signal extraction from noise has been traditionally solved by registering a great number of individual trials and averaging the values obtained on each "t" point of the time series. This proceeding, commonly used on most laboratories, presents serious application problems on the CNV complex. Following Tukey's suggestions (1978), it is proposed on the thesis the use, along with averaging, of some smoothing technique (Tukey, 1977; Velleman & Hoaglin, 1981). These techniques application yields to a perspective change on the signal estimation, by substituting classic transversal averaging that uses simple averaging, by a longitudinal one.

In the thesis it has been done an application of smoothing techniques to a data obtained on CNV and RP (Readiness Potential) experiments. On both, combined application of smoothing and simple averaging offers an signal estimation from a reduced number of individual trials, avoiding in such a way the fatigue and habituation processes. Is this one of the major advantages comparing to the simple averaging that needs a great number of trials to obtain a similar estimation.

Il progresso tecnologico e la riduzione dei costi ha significativamente contribuito all'identificazioni delle cause genetiche di diverse malattie genetiche multifattoriali. In questo lavoro di dottorato riporto i risultati di un analisi genetica di una coorte di ragazzi dislessici utilizzando delle tecnologie ad alta processività come Next Generation Sequencing e SNP-array ad alta densità. La coorte in studio consiste di 49 soggetti con dislessia e 52 soggetti con dislessia e altre Disabilità Specifiche di Apprendimento (disortografia, disgrafia, discalculia). Tutti i campioni sono stati sequenziati utilizzando la piattaforma Ion Torrent, focalizzandosi sulle regioni codificanti e sulle giunzioni esone-introne, in 12 geni candidati (CMIP, CNTNAP2, CYP19A1, DCDC2, DIP2A, DYX1C1, GCFC2, KIAA0319, KIAA0319L, MRPL19, ROBO1, S100B), focalizzandosi sulle varianti non descritte in letteratura e in quelle rare (MAF <1%). Un sotto-gruppo di 54 soggetti è stato ulteriormente analizzato, genotipizzando più di 1,7 M marcatori (Multi Ethnic Global Array design, Ilumina), per l'identificazione e caratterizzazione in base alla letterature di Copy Number Variation (CNV). Per questo proposito, per la chiamata di CNV ad "alta fedeltà", sono stati usati due algoritmi: cnvPartition e PennCNV. In totale riportiamo 12 varianti predette patogenetiche, tra le qualli due nuove con effetto deleterio (DIP2A:p.G1387* and KIAA0319:p.V774Afs*37) e una rara variante di splicing (GCFC2:c.266-2A>G). Inoltre, diverse CNV sono state identificate che sovrappongono dei geni associati a problemi con il linguaggio, ma nessuna delle CNV con i geni riportati sopra. Infine, una copia di fratelli sono portatori di diverse duplicazioni localizzate nella regione 16p13.11, una regione di suscettibilità ai disordini di neurosviluppo. Il presente lavoro arrichisse la nostra conoscenza sul background genetico della coorte in studio. Nello stesso momento i risultati ottenuti devono essere ulteriormente analizzati per poter attribuire un ruolo a quanto possibile certo sul loro contributo all'insorgenza della dislessia.
Technological improvements and continued cost reduction have significantly contributed to the progress of identifying the genetic causes of complex traits. Here we report the results of a genetic screening on a dyslexia cohort combining targeted next generation sequencing and high density SNP array. The study cohort consists of 49 subjects with dyslexia and 52 subjects with dyslexia and other specific learning disabilities (dysorthographia, dysgraphia, dyscalculia). All samples were sequenced on Ion Torrent platform, targeting the coding regions and their exone-intron boundaries of 12 candidate dyslexia genes (CMIP, CNTNAP2, CYP19A1, DCDC2, DIP2A, DYX1C1, GCFC2, KIAA0319, KIAA0319L, MRPL19, ROBO1, S100B), with focus on novel and rare variants. A subset of 54 samples was further analyzed, genotyping over 1.7 M markers (Multi Ethnic Global Array design, Ilumina), for copy number variation (CNV) discovery and characterization according to the literature. For this purpose, high confidence CNVs were obtained using the cnvPartition and the PennCNV calling algorithms. We report a total of 12 pathogenic predicted variants, among which two novel deleterious events (DIP2A:p.G1387* and KIAA0319:p.V774Afs*37) and a known rare splicing variant (GCFC2:c.266-2A>G). Moreover, several copy number variants were identified, overlapping some language related genes, but not any of the above sequenced genes. Finally, a sibling pair was found to harbor duplications in the chromosome band 16p13.11, a susceptibility region for several neurodevelopmental disorders. The present study enriches our knowledge about the genetic background in a dyslexia cohort. At the same time our findings emphasize the need for further research to attribute causative roles of these events for cohort phenotypes.

Deficiência intelectual é uma condição heterogênea e complexa, diagnosticada em 1-3% da população mundial. Desequilíbrios cromossômicos e variações no número de cópias (CNVs) são as causas mais frequentes de DI e, até recentemente, a maior parte desse desequilíbrio era averiguado por análises citogenéticas convencionais. Antes da utilização de microarrays cromossômicos (CMA), a causa etiológica da DI ainda permanecia desconhecida em ~60% dos pacientes. A aplicação de CMA tem revolucionado o diagnóstico da DI e de muitas outras doenças congênitas, permitindo explicar a etiologia molecular de parte da DI através da identificação de CNVs patogênicas. Nos países desenvolvidos, CMA é considerado como primeiro teste para avaliar pacientes com múltiplas anomalias congênitas, DI e/ou autismo. Contudo, nos países em desenvolvimento, a detecção de alterações ainda é feita principalmente por métodos citogenéticos convencionais. O objetivo desse estudo foi identificar, através do uso de SNP arrays, o espectro de anomalias cromossômicas presente em uma amostra de 40 pacientes com DI idiopática moderada e grave, apresentando ou não aspectos dismórficos e anomalias congênitas. Em especial, essa coorte de pacientes, em sua maioria (~2/3), não havia sido previamente cariotipada. Embora mundialmente desde 2010 a recomendação seja de realizar arrays antes de cariótipo, a maioria dos pacientes relatados em estudos já havia sido cariotipada antes de array ser oferecido a eles como teste. Foram identificadas alterações raras em 18 pacientes (45%). Em 12 (30%) desses Pacientes, as CNVs eram sabidamente patogênicas; esta taxa diagnóstica está muito acima da taxa de detecção reportada na literatura (~20%) e possíveis causas desta discrepância são discutidas. Outros 6 Pacientes (15%) apresentaram variantes raras de significado incerto (variants of unknown significance - VUS). Um aspecto adicional investigado foram os mecanismos envolvidos na formação de alguns dos rearranjos estruturais; enquanto nosso foco inicial era o uso de arrays para detecção de CNVs, se tornou evidente no decorrer do projeto que o padrão dos SNPs obtido nos arrays revelava, a partir do DNA, informação valiosa sobre a estrutura dos cromossomos e a composição heterogênea de células em uma amostra (mosaicismo). Esses resultados são discutidos em detalhes em duas situações: (1) A descrição de uma deleção terminal 1p36, associada a dissomia uniparental (UPD) em mosaico de segmentos de 1pter de diferentes tamanhos. Sugerimos que essa composição reflita eventos recorrentes de captura de telômero, embora processo similar nunca tenha sido descrito, e propomos um possível mecanismo responsável por originar esse desequilíbrio complexo. (2) Três dos nossos pacientes apresentam 4 cópias ou uma combinação de 3-4 cópias de segmentos proximais, na maior parte superpostos, de 15q11q13. Possíveis mecanismos de origem desses rearranjos são discutidos
Intellectual disability (ID) is a complex and heterogeneous condition affecting about 1-3% of the general population. Chromosomal imbalances and copy-number variations (CNVs) have been recognized as the most frequent causes of ID and, until recently, most of these imbalances were diagnosed by cytogenetic analysis. Before the application of microarray analysis (CMA), the underlying cause of ID remains unknown in ~60% of patients. The use of CMA has revolutionized the diagnosis of ID and several other congenital disorders, and have made it possible to identify pathogenic CNVs that could explain the molecular etiology of ID. In developed countries, CMA is considered the first-tier technique for the analysis of patients with multiple congenital anomalies, ID, and/or autism spectrum disorders. However, in developing nations, detection of alterations is still performed mainly by conventional cytogenetic techniques. The aim of this study was identifying, using a high-density resolution SNP microarray, chromosomal imbalances in a total of 40 patients presented with moderate-to-severe ID, associated or not with dysmorphic features and congenital anomalies. Particularly, most of the patients in the cohort (~2/3) was not karyotyped previously. Although CMA has been recommended as the first-tier test since 2010 all over the world, the majority of the patients in the reported studies were karyotyped before CMA was offered as a diagnostic test. Rare CNVs were detected in 18 patients (45%). Among those patients, 12 (30%) carried pathogenic CNVs. This yield is much higher than reported in the literature (~20%), and possible causes for this discrepancy are discussed. Six patients (15%) carried variant of unknown significance (VUS). Furthermore, mechanisms involved in structural rearrangements found in some patients were investigated. Even though the main focus of this dissertation was the detection of CNVs using high resolution SNP arrays, throughout the course of this project it was clear that the SNP patterns found could reveal crucial information about the structure of chromosomes and the heterogeneous composition of cells (mosaicism). Those results are discussed in detail in two situations: (1) One description of a terminal 1p36 deletion, associated with mosaic uniparental disomy (UPD) of different sized 1pter segments. We hypothesized that this composition reflects recurrent telomere capture events, although a similar process has never been described so far, and proposed a possible mechanism responsible for originating this complex imbalance. (2) Three of our patients carried four copies or a four-three copies-combination of a proximal, partially overlapping, 15q11q13 segment. Possible mechanisms responsible for this complex rearrangement are discussed

Copy number variation (CNV) is a relatively novel source of genetic variation, involving the duplication or deletion of segments of genomic DNA (gDNA) sequence, thereby changing the original number of DNA copies. It is currently gaining widespread recognition from the scientific community, and it is anticipated to play a major role in the aetiology of human diseases. However, the extent of its contribution to phenotypic diversity, in terms of individual susceptibility to disease, remains to be elucidated. Nonetheless, recent studies have indicated that common complex disease phenotypes, such as osteoporosis, might be highly susceptible to CNV. Osteoporosis is a common and debilitating skeletal condition, imposing significant clinical and socioeconomic consequences. The disease is characterised by fragile bones that are susceptible to fracture due to deregulated bone remodelling, where bone loss exceeds bone formation. Being a common complex disease, osteoporosis risk is largely determined by the effect of environmental factors on genetic variants. Moreover, identification of the genetic variants associated with osteoporosis is widely anticipated for the contribution it will make towards the development of improved measures of disease intervention. Recent genome-wide association studies (GWASs) have identified that the genes oestrogen receptor 1 (ESR1) and Axin 1 (AXIN1) potentially play major roles in bone regulation. In addition, evidence highlights their involvement in key biological processes that regulate bone turnover. Specifically, ESR1 mediates the response of bone marrow-derived cells to oestrogen and it has been demonstrated that oestrogen inhibits bone loss, while AXIN1 inhibits Wnt signal transduction and it has been demonstrated that Wnt proteins promote bone growth. Furthermore, several large-scale analysis projects firmly implicate genetic variations of both genes with bone marrow density (BMD), which is the surrogate phenotype of osteoporosis. Therefore, ESR1 and AXIN1 are both recognised candidates for the genetic regulation of osteoporosis risk. This study investigated the potential effect of two novel CNVs of the genes ESR1 and AXIN1, Variant_4512 and Variant_4912, respectively, in relation to BMD in a population cohort study of Caucasian women, between the ages of 18 and 83, from Australia and the UK. Subjects were genotyped for both CNVs, respectively, using real-time quantitative PCR (qPCR) combined with TaqMan chemistry, and the copy number (CN) quantitation software, CopyCaller. Subjects were then examined for evidence of association between both CNVs and three different BMD phenotypes, 1) raw measurement (g/cm2), 2) age-adjusted Z-score, and 3) controlled for several covariates, at three common skeletal locations of osteoporotic fracture, 1) lumbar spine, 2) total hip, and 3) femoral neck. This study confirmed the presence of ESR1 CNV and AXIN1 CNV in the analysed subject cohort, as indicated by the observation of three distinct CNV genotypes for each, representing CN loss (CN1) and CN gain (CN3) from the expected wild-type CN in the human diploid genome (CN2). This study found no evidence of association between both CNVs and BMD (p = > 0.05) in the analysed subject cohort. Therefore, the hypothesis tested in this study, that CNV is associated with BMD, was not supported. As a result, it would appear that the ESR1 CNV Variant_4512 and the AXIN1 CNV Variant_4912 are unlikely to play a major role in the pathogenesis of osteoporosis in Caucasian women. However, replication studies and further research would be required to accurately validate this, since this study was subject to numerous limitations which may have influenced the findings, such as low statistical power, technical difficulties, limiting experimental reagents, and time constraints. In addition, there is evidence from previous studies implicating intron 1 and the 5’ end of ESR1 and intron 2 of AXIN1 with BMD. Variant_4512 and Variant_4912 encompass the 5’ end of their respective genes, thereby implicating the promoter sequence and regulatory elements, which in turn implicates the control of gene expression. Therefore, despite the lack of statistically significant findings in this study, the ESR1 CNV Variant_4512 and the AXIN1 CNV Variant_4912 both still remain as promising candidates for involvement in BMD and the risk of osteoporosis. Moreover, other CNVs in the same genomic regions may also be relevant for future research. Further research would benefit from addressing the potential effect of environmental risk factors on CNV. It is possible that the ESR1 CNV Variant_4512 may be modified in an environment-specific manner, which influences its effect on BMD, as indicated by the almost statistically significant association between ESR1 CNV and BMD observed in this study when controlled for covariates at the femoral neck (p = 0.052). Moreover, previous studies highlight that the majority of known genes subject to CNV are not even located within the identified region of genomic variation, and also that osteoporosis may be more susceptible to genetic variation affecting the CN of non-coding regions. Therefore, further research should also focus on gene expression studies to determine whether the ESR1 CNV Variant_4512 exerts position effects on the transcriptional control of another gene, which may in turn be the primary gene associated with osteoporosis.

The 16p11.2 copy-number variant (CNV) represents a well-characterized, high-risk factor for autism spectrum disorder that additionally predisposes deletion carriers (16pdel) to increased head circumference, known as macrocephaly. The 16p11.2 CNV consists of 29 known genes, many of which are associated with neurobiological processes relevant for macrocephaly such as cell proliferation and apoptosis, differentiation and cell growth. Our lab’s previous work has demonstrated that induced pluripotent stem cell (iPSC)-derived neurons from 16pdel carriers show altered cellular morphology related to growth, which include increased soma size, total dendritic length and dendritic complexity. However, specific CNV genes responsible for these phenotypes have not been established. Here, we investigate the relationship between three 16p11.2 genes and the observed cellular phenotypes. We differentiated neurons from control iPSC-derived neural progenitor cells (NPCs) and used short hairpin RNA (shRNA) to reduce the expression of these CNV genes: KCTD13, MAPK3 and C16ORF53. We then assessed neuronal morphology by evaluating soma size, total dendritic length and dendritic complexity. We demonstrate that knocking down KCTD13 and C16ORF53 increases soma size and total dendrite length, respectively, similar to that observed in 16pdel iPSC-derived neurons. For this reason, we speculate that these genes may have a role in cell growth and might underlie macrocephaly. Thus, our study investigates genes in the 16p11.2 CNV that contribute to neuronal morphology, which may have a role in influencing brain size.

Submitted by admin tede (tede@pucgoias.edu.br) on 2018-06-19T18:06:59Z No. of bitstreams: 1 SAMARA SOCORRO SILVA PEREIRA.pdf: 764936 bytes, checksum: 434a8cd13e6701c05b6ea08edda95150 (MD5)
Made available in DSpace on 2018-06-19T18:06:59Z (GMT). No. of bitstreams: 1 SAMARA SOCORRO SILVA PEREIRA.pdf: 764936 bytes, checksum: 434a8cd13e6701c05b6ea08edda95150 (MD5) Previous issue date: 2018-03-14
Copy Number Variation (CNV) has been associated with intelectual disability (ID) and this condition occur in approximately 2% of world population. Chromosomal Microarray Analysis (CMA) is being indicated as first-tier test for individuals with ID and has also helped to understand the mechanisms of CNV formation, classification of these rearrangements, type of recurrence, and its origin. The aim of this study was to infer the parental chromosome origin of de novo pathogenic CNV in patients with ID and their mechanisms of formation. Patients with clinical indications of ID were referred to Replicon Research Group/LaGene for G-band karyotyping. CMA approach was done for patients without numerical and/or structural rearrangements results in karyotype. After performing CMA and classification of CNVs, the parental origin of pathogenic CNVs was done using Mendelian error check based on SNPs markers available by ChAs software. In addition, the UCSC Genome Browser website was used to detect Low Copy Repeats (LCR) surrounding the CNVs to infer the mechanisms of their formation. In the period from 2013 to 2015 was performed G-band karyotyping in 290 patients with clinical indication of ID and a total of 193/290 (66.5%) were diagnosed by Karyotype. The group of patients who were not diagnosed using the karyotype, only 76/97 (78.3%) agreed to continue the investigation by CMA’s approach. After performing CMA, a total of 15 de novo pathogenic CNVs were observed, 10 CNV of loss and 5 CNV of gain, in 13/76 (17.1%) patients. The analysis of the parental origin showed 60% of CNVs are of maternal origin and 40% of paternal origin. It was not possible to detect the influence of parental age in the formation of CNVs. After analyzing the presence of surrounding LCRs, it was observed that 46.7% are recurrent CNVs and the mechanism of formation was Non- Allelic Homologous Recombination (NAHR), and 71.4% of these recurrent CNVs are of maternal origin. These data are in agreement with studies that affirm that the majority of CNVs of paternal origin are nonrecurrent due to germ cells replicate many times their genetic material in the pre-meiotic phase, being possible to infer the mechanism of formation of CNV that may have been by Microhomology-mediated break-induced replication (MMBIR) or Non-homologous end joining (NHEJ).
A variação no número de cópias (CNV) no genoma é um dos fatores etiológicos que pode desencadear a condição da deficiência intelectual (DI), sendo que esta condição atinge cerca de 2% da população mundial. A metodologia de análise cromossômica por microarranjo (CMA) além de ser indicada como teste de primeira escolha para pacientes com DI, tem ajudado também na compreensão da formação de CNVs e classificação destes rearranjos, quanto à patogenicidade, o tipo de recorrência e sua origem. E este estudo objetivou inferir a origem cromossômica parental das CNVs de novo patogênicas em pacientes com DI e seu mecanismo de formação. Os pacientes com indicação clínica de DI foram encaminhados ao Núcleo de Pesquisas Replicon/LaGene para realização do cariótipo com bandeamento GTG, e subsequentemente, os que não tiveram alteração numérica e/ou estrutural no cariótipo foram convidados a continuar a investigação em nível genômico, pela metodologia de CMA. Após realização do CMA e classificação das CNVs, foram realizadas a análise da origem parental das CNVs de novo patogênicas pela análise do erro mendeliano usando os marcadores de SNPs disponibilizado pelo software ChAS. Adicionalmente, foi usado o UCSC Genome Browser para detectar Repetições De Poucas Cópias (LCR) circundantes as CNVs para inferir o mecanismo de formação das mesmas. Foi realizado o cariótipo em 290 pacientes com indicação clínica de DI entre os anos de 2013 a 2015 e em 193/290 (66,5%) foram diagnosticados pelo cariótipo. Do conjunto de pacientes que não foram diagnosticados usando o cariótipo, apenas 76/97 (78,3%) aceitaram continuar a investigação pelo CMA. Após realizar o CMA, foi observado 15 CNVs de novo patogênicas, 10 CNVs de perda e 5 CNVs de ganho, em 13/76 (17,1%) pacientes. Na análise da origem parental, observou-se que 60% das CNVs são de origem materna e 40% de origem paterna. Não foi possível detectar a influência da idade parental na formação das CNVs. Ao analisar a presença de LCRs circundantes, observou-se que 46,7% das CNVs de novo patogênicas são recorrentes e o mecanismo de formação foi a Recombinação Homologa Não Alélica (NAHR), e 71,4% dessas CNVs recorrentes são de origem materna. Esses dados corroboram com os estudos que afirmam que a maioria das CNVs de origem paterna são não recorrentes devido às células germinativas replicarem inúmeras vezes o seu material genético na fase pré-meiótica, sendo possível inferir sobre o mecanismo de formação que pode ter sido por Replicação Induzida por Quebra e Mediada por Microhomologia (MMBIR) ou Junção de Extremidade Não Alélica (NHEJ).

Microduplication of 16p11.2 (dup16p11.2) is associated with a broad spectrum of neurodevelopmental disorders (NDD) confounded by variable expressivity. I hypothesized that while some unique features reported in individuals with dup16p11.2 may be explained by the over-expression of its integral genes, co-occurrence of other genetic alterations in the genome may account for the variability in their clinical phenotypes. This hypothesis was explored in two unrelated subjects with NDD who each inherited the dup16p11.2 from an apparently healthy carrier parent. First, I performed a detailed phenotypic analysis of individuals with dup16p11.2 (published and current study). I did not find evidence of phenotypic commonality and consistent syndromic phenotype pattern among carriers of dup16p11.2. Next, I assessed the effect of dup16p11.2 on the expression of genes located within and nearby this region which showed that RNA expression of KCTD13, MAPK3 and MVP from 16p11.2 region was inconsistent in carriers of dup16p11.2. KTCD13 has been identified as a driver of a mirror brain phenotype of 16p11.2 CNV in zebrafish. However, the data presented here demonstrated that dup16p11.2 did not result in increased protein expression of KCTD13 in either probands or the healthy carrier parent, indicating that KCTD13 is not the sole cause of microcephaly in cases with dup16p11.2. Finally, whole exome sequencing (WES) was used to investigate the presence of genomic sequence changes in dup16p11.2 carriers that could explain such clinical variability. Compound heterozygous variants of VPS13B in proband A and missense variants of SYNE2 in proband B were identified. Mutations of VPS13B cause Cohen syndrome in keeping with proband A’s phenotype (ID, microcephaly, facial gestalt, retinal dystrophy, joint hypermobility and episodic neutropenia) and low RNA expression. The protein encoded by SYNE2, Nesprin 2, plays critical roles in neurogenesis in mice. Over-expression of Nesprin 2 identified in proband B may cause NDD phenotypes via a dominant-negative effect. In conclusion, pathogenic variants were identified in genes outside of the 16p11.2 region which could contribute to the clinical variability between parent-offspring dup16p11.2 carriers in this study. This suggests discordance in phenotype of dup16p11.2 carriers warrants further study by WES and individualized genetic assessment and counselling.
Medicine, Faculty of
Medical Genetics, Department of
Graduate

A Thesis submitted to The University of Arizona College of Medicine - Phoenix in partial fulfillment of the requirements for the Degree of Doctor of Medicine.
Background: Bipolar disorder (BD) is a disabling disorder whereby individuals suffer from episodes of mania and depression. The mode of inheritance of BD is complex and likely multifactorial. The specific number of susceptibility loci, the recurrence risk ratio attributable to each locus, and the degree of interaction between loci are unknown. By determining whether single nucleotide polymorphisms (SNPs) or copy number variants (CNVs) predispose individuals to bipolar disorder, therapeutics and diagnostic tests may be developed. Method: A Genome Wide Association Study (GWAS) was performed using cases of bipolar disorder and normal controls hybridized on Affymetrix 6.0 Genome-Wide Human SNP Arrays. Data preprocessing removed 595 individuals from 2205 arrays. The probe intensities of the remaining 880 cases and 730 controls were normalized. A modified t-test algorithm was used to determine p-values for each SNP. A sliding window analysis was performed on SNPs ordered by chromosome and locus. The mean probe intensities of the cases and controls from regions of significance were then reanalyzed for differences. Results: Analysis yielded several SNPs and CNVs that may have involvement in the pathophysiology of bipolar disorder. One region was 15kbp within the Neuron Navigator 2 (NAV2) gene. A second region was found in the Down Syndrome Cell Adhesion Molecule Like 1 (DSCAML1) gene. A third region was within the Voltage-dependent Calcium Channel Alpha 1G (CACNA1G). Conclusion: Multiple SNPs and CNVs may play a role in the phenotype of Bipolar Disorder. A convergent functional genomics approach with a gene network analysis maybe warranted elucidating possible pathophysiologies involving the gene products found to be significant in this study.

Sintesi – Italiano Gli scopi, i materiali, i metodi usati, i risultati e le conclusioni dei tre studi sono organizzati in tre capitoli. La sintesi generale dei tre studi e quindi divisa in base a questi tre capitoli. Capitolo 1 La determinazione dei “copy number variants” (CNV) è fondamentale per la valutazione dei tratti genomici in diverse specie in quanto rappresentano una fonte principale della variabilità genetica, influenzando l’espressione genica, la variabilità fenotipica, la adattabilità e la predisposizione all’insorgenza di malattie. Lo scopo di questo studio è stato quello di ottenere una mappa genomica di CNV utilizzando i dati ottenuti dall’Illumina Bovine SNP50 BeadChip di 651 tori di razza Bruna Italiana. Per l’identificazione dei CNV e delle regioni CNV (CNVR) sono stati usati i software PennCNV e SVS7 (Golden Helix). Sono stati identificati un totale di 5,099 e 1,289 CNVs con i software PennCNV ed SVS7 rispettivamente. Questi CNV sono stati raggruppati a livello di popolazione in 1,101 (220 delezioni, 774 duplicazioni e 107 complex) e 277 (185 delezioni, 56 duplicazioni e 36 complex) CNVR. Dieci dei CNVR selezionati sono stati validati sperimentalmente attraverso qPCR. La GO e la pathway analysis effettuate hanno identificato i geni (corretti per la false discovery rate) localizzati nelle CNVR e correlati a diversi processi biologici, componenti cellulari, funzioni metaboliche e vie metaboliche. Tra questi, sono stati identificati i geni FCGR2B, PPARalpha, KATNAL1, DNAJC15, PTK2, TG, STAT family, NPM1, GATA2, LMF1 e ECHS1, già noti in letteratura, per la loro associazione con diversi caratteri quantitativi nei bovinia. Sebbene ci sia una variabilità nell’identificazione dei CNVR attraverso l’utilizzo di diversi metodi e piattaforme, questo studio ha permesso l’identificazione dei CNVR nella Bruna Italiana, sovrapponendo quelli già identificati in altre razze e identificandone dei nuovi, producendo quindi nuove conoscenze per gli studi di associazione con caratteri quantitativi di interesse nei bovini. Capitolo 2 Scoprire variazioni genetiche come i Copy Number Variants (CNVs) nei bovini, fornisce l’opportinità di studiare la loro associazione con caratteri quantitativi. I CNVs sono sequenze di DNA di lunghezza 50 bp fino a diverse Mb, che possono variare in numero di copie rispetto ad un genoma di riferimento. Lo scopo di questo studio è stato quello di identificare i CNVs in 1,410 campioni di razza Bruna Svizzera usando informazioni derivanti dall’ Illumina Bovine HD SNP chip, che include 777,962 SNPs. Dopo uno stringente controllo di qualità, i CNVs sono stati identificati con i software Golden Helix SVS 8.3.1 (SVS) e PennCNV e sono stati raggruppati in regioni CNV (CNVRs) a livello di popolazione (i.e. CNVs sovrapposti) utilizzando il software BEDTools. I CNVR comuni ai due software sono stati definiti come regioni consensus. I geni all’interno delle CNVR consensus sono stati annotati con un’analisi GO utilizzando DAVID Bioinformatics Resources 6.7. Per poter validare i risultati, sono state eseguite PCR quantitative su 15 CNVR selezionate. Con il software SVS sono stati identificati 25,030 CNVs successivamente raggruppati in 398 CNVR, che comprendevano 30 duplicazioni, 344 delezioni e 24 complex CNVR (che contenevano sia duplicazioni che delezioni) coprendo il 3.92% del genoma bovino. Il software PennCNV ha identificato 62,341 CNV, corrispondenti a 5,578 CNVRs che comprendevano 2,638 duplicazioni, 2,404 delezioni e 537 complex CNVR, coprendo il 7.68% del genoma bovino. La lunghezza di queste CNVR variava da 1,244 bp a 1,381,355 bp. Sono state trovate 563 CNVR consensus che coprivano il 2.29% del UMD 3.1 bovine genome assembly. Di queste, 24 erano duplicazioni, 300 erano delezioni e 239 erano CNVR complex. Un totale di 775 official gene IDs sono stati annotati nelle CNVR consensus. Tra i 537 geni con informazioni funzionali, la GO e la pathway analysis è stata riportata per quelli che clusterizzavano con un p-value < 0.05. Le PCR quantitative hanno validato con successo 14 delle 15 CNVR selezionate. Il risultato di questo studio è una prima analisi genomica integrale della razza Bruna Svizzera basata sull’Illumina Bovine HD SNP chip su un numero cosi grande di animali che arricchisce la mappa CNV nel genoma bovino. I risultati forniscono inoltre informazioni preziose per successivi studi sui CNV. Infine, i risultati della mappa CNVR sono informativi per i caratteri funzionali, produttivi e sanitari considerati nei programmi di selezione nella razza Bruna Svizzera. Capitolo 3 I Copy Number Variations (CNV) possono essere usati negli studi di associazione per rivelare la base genetica della variazione fenotipica di caratteri quantitativi. I CNV sono sequenze di DNA di 50 bp fino a qualche Mb, che possono variare in numero di copie rispetto ad un genoma di riferimento. Fino ad oggi, nessuno studio di associazione genome-wide (GWAS) con i CNV e caratteri quantitavivi è stato descritto in una popolazione cosi ampia (cioè di 1,116 campioni) della razza bovina Bruna Svizzera. Lo scopo di questo studio era quello di eseguire delle GWAS utilizzando i CNV precedentemente mappati, con caratteri funzionali, produttivi e sanitari al fine di valutare il loro impatto sull’allevamento e sulla selezione. Gli studi di associazione con i CNV sono stati effettuati con il software Golden Helix SVS 8.4.4 utilizzando un correlation-trend test model. I geni all’interno dei CNV significativamente associati per ogni carattere sono stati annotati con un’analisi GO usando DAVID Bioinformatics Resources 6.7. Sono stati identificati 56 CNV significativamente associati con uno o più degli otto caratteri valutati. I segnali di associazione più forti erano dati da tre CNV sul cromosoma 12 per il carattere grasso. I CNV associati si sovrappongono con 23 geni diversi, annotati sul Bos taurus genome assembly (UMD3.1).
Abstract – English The aims, material and methods, results and conclusions of the three studies are organized in three different chapters. The general abstract is therefore divided according to these chapters. Chapter 1 The determination of copy number variation (CNV) is very important for the evaluation of genomic traits in several species because they are a major source for the genetic variation, influencing gene expression, phenotypic variation, adaptation and the development of diseases. The aim of this study was to obtain a CNV genome map using the Illumina Bovine SNP50 BeadChip data of 651 bulls of the Italian Brown Swiss breed. PennCNV and SVS7 (Golden Helix) software were used for the detection of the CNVs and Copy Number Variation Regions (CNVRs). A total of 5,099 and 1,289 CNVs were identified with PennCNV and SVS7 software, respectively. These were grouped at the population level into 1101 (220losses, 774 gains, 107 complex) and 277 (185losses, 56 gains and 36 complex) CNVRs. Ten of the selected CNVRs were experimentally validated with a qPCR experiment. The GO and pathway analyses were conducted and they identified genes (false discovery rate corrected) in the CNVR related to biological processes, cellular component, molecular function and metabolic pathways. Among those, we found the FCGR2B, PPARalpha, KATNAL1, DNAJC15, PTK2, TG, STAT family, NPM1, GATA2, LMF1, ECHS1 genes, already known in literature because of their association with various traits in cattle. Although there is variability in the CNVRs detection across methods and platforms, this study allowed the identification of CNVRs in Italian Brown Swiss, overlapping those already detected in other breeds and finding additional ones, thus producing new knowledge for association studies with traits of interest in cattle. Chapter 2 Detecting genetic variation such as Copy Number Variants (CNVs) in cattle provides the opportunity to study their association with quantitative traits. CNVs are DNA sequences of 50 bp up to several Mb long, which can vary in copy number in comparison with a reference genome. The aim of this study was to investigate CNVs in 1,410 samples of the Brown Swiss cattle breed using Illumina Bovine HD SNP chip information, which includes 777,962 SNPs. After stringent quality control, CNVs were called with the Golden Helix SVS 8.3.1 (SVS) and PennCNV software and were summarized to CNV regions (CNVRs) at a population level (i.e. overlapping CNVs), using BEDTools. Additionally, common CNVRs between the two software were set as consensus regions. Genes within consensus CNVRs were annotated with a GO analysis using the DAVID Bioinformatics Resources 6.7. In order to validate these results, quantitative PCRs were executed on 15 selected CNVRs. The SVS software identified 25,030 CNVs summarized to 398 CNVRs, which comprised 30 gains, 344 losses and 24 complex CNVRs (i.e. containing both losses and gains), covering 3.92% of the bovine genome. The PennCNV software identified 6,2341 CNVs summarized to 5,578 CNVRs, which comprised 2,638 gains, 2,404 losses and 537 complex CNVRs, covering 7.68% of the bovine genome. The length of these CNVRs ranged from 1,244 bp to 1,381,355 bp. A total of 563 consensus CNVRs were found covering 2.29 % of the UMD 3.1 bovine genome assembly. Of these, 24 were gains, 300 were losses and 239 were complex CNVRs. A total of 775 official gene IDs were annotated in the consensus CNVRs. Among the 537 genes with functional information, the GO and pathway analysis was reported for those who clustered with a p-value < 0.05. The quantitative PCRs successfully validated 14 (93.33%) of the selected CNVRs. The result of this study is the first comprehensive genomic analysis of the Brown Swiss breed based on the Illumina Bovine HD SNP chip on such a large number of animals that enriches the CNV map in the bovine genome. These findings also provide valuable information for further CNV studies. Finally, the results of the CNVR map delivers new information for functional, health and productive traits considered in selection programs of the Brown Swiss breed. Chapter 3 Copy Number Variation (CNV) can be used in association studies to disclose genetic basis of quantitative traits phenotypic variation. CNVs are DNA sequences of 50 bp up to several Mb long, which can vary in number of copies in comparison with a reference genome. Up to date, no genome-wide association study (GWAS) with CNVs and quantitative traits in such a large Brown Swiss population (i.e. with 1,116 samples) has been described. The purpose of this study was to perform a GWAS using CNVs with functional, health and productive traits and to asses the impact on farming and breeding practices. The CNV – association studies were performed with the Golden Helix SVS 8.4.4 software using a correlation-trend test model. Genes within significant associated CNVs for each trait were annotated with a GO analysis using the DAVID Bioinformatics Resources 6.7. A total of 56 CNVs were significantly associated with one or more of the eight evaluated traits. The greatest association signals were given by three CNVs on chromosome 12 for the fat yield trait and on BTA23 for udder traits. The associated CNVs overlap with 23 different genes annotated on the Bos taurus genome assembly (UMD3.1).

Copy number Variants (CNVs), which comprise deletions, insertions and inversions of genomic sequence, are a main form of genetic variation between individual genomes. CNVs are commonly present in the genomes of human and other species. However, they have not been extensively characterized as their ascertainment is challenging. I reviewed current CNV studies and CNV discovery methods, especially the algorithms which infer CNVs from whole genome Single Nucleotide Polymorphism (SNP) arrays and compared the performance of three analytical tools in order to identify the best method of CNV identification. Then I applied this method to identify CNV events in three European population isolates—the island of Vis in Croatia, the islands of Orkney in Scotland and villages in the South Tyrol in Italy - from Illumina genome-wide array data with more than 300,000 SNPs. I analyzed and compared CNV features across these three populations, including CNV frequencies, genome distribution, gene content, segmental duplication overlap and GC content. With the pedigree information for each population, I investigated the inheritance and segregation of CNVs in families. I also looked at association between CNVs and quantitative traits measured in the study samples. CNVs were widely found in study samples and reference genomes. Discrepancies were found between sets of CNVs called by different analytical tools. I detected 4016 CNVs in 1964 individuals, out of a total of 2789 participants from the three population isolates, which clustered into 743 copy number variable regions (CNVRs). Features of these CVNRs, including frequency and distribution, were compared and were shown to differ significantly between the Orcadian, South Tyrolean and Dalmatian population samples. Consistent with the inference that this indicated population-specific CNVR identity and origin, it was also demonstrated that CNV variation within each population can be used to measure genetic relatedness. Finally, I discovered that individuals who had extreme values of some metabolic traits possessed rare CNVs which overlapped with known genes more often than in individuals with moderate trait values.

Le raisonnement conditionnel, fondé sur les énoncés de la forme Si P alors Q, est celui qui a reçu le plus d'attention de la part des psychologues. Les arguments principaux du raisonnement conditionnel, comme le Modus Ponens (MP), sont constitués de trois éléments : la prémisse majeure (Si P alors Q), la prémisse mineure (P) et la conclusion (Q). Ces éléments constituent trois étapes de traitement distinctes. Cependant, la dimension temporelle du raisonnement a été en partie négligée dans la littérature. L’objectif central de cette thèse a été d’explorer cette dimension temporelle à l’aide d’une approche novatrice combinant l’utilisation de la mesure du temps de lecture des prémisses, de l’Electroencéphalographie (EEG) et de la Magnétoencéphalographie (MEG). Nous nous sommes donné trois objectifs : 1) Déterminer la séquence des étapes de traitement élémentaire de l’argument MP ; 2) Déterminer comment l’incertitude d’un conditionnel thématique est prise en compte ; 3) Mettre en évidence les différences interindividuelles de traitement d’un énoncé conditionnel, basique ou thématique, en introduisant l’étude de l’argument AC qui permet de dissocier deux populations : les individus qui acceptent la conclusion de AC et les individus qui la rejettent.L’ensemble des données révèle que tous les individus ont une tendance à se focaliser davantage sur P que sur Q lors du traitement du conditionnel, avec des degrés variables selon les individus. Lorsque la prémisse P (ou Q pour les participants qui acceptent AC) est présentée, elle est intégrée à la prémisse majeure afin de générer une conclusion Q encodée et stockée en mémoire de travail avant d’être comparée avec la conclusion présentée.Lorsque le conditionnel est incertain (conditionnel thématique), cette incertitude sur la suffisance de P pour Q (ou de Q pour P) semble être prise en compte par les sujets au niveau de la prémisse majeure et se manifeste par une attente moins prononcée de la conclusion Q une fois que la prémisse P a été présentée
The conditional reasoning, based on statements of the form If P then Q, is one which has received the most attention from psychologists. The main arguments of conditional reasoning, as the Modus Ponens (MP), consist of three elements: the major premise (If P then Q), the minor premise (P) and conclusion (Q). These elements constitute three separate processing steps. However, the temporal dimension of reasoning has been partly neglected in the literature. The central objective of this thesis was to explore the temporal dimension by using an innovative approach combining the use of the measurement of premises reading time and of the electroencephalography (EEG) and magnetoencephalography ( MEG). We set three objectives: 1) Determine the sequence of processing steps of the basic argument MP 2) Determine how the uncertainty of a conditional theme is taken into account, 3) Highlight the interindividual differences in treatment a conditional statement, or basic theme by introducing the study of the AC argument, which allows to separate two populations: individuals who accept the conclusion of AC and individuals who reject it. The data reveals that all individuals have a tendency to focus more on P and Q in the processing of the conditional, with varying degrees in different individuals. When the premise P (or Q for participants that accept AC) is presented, it is integrated with the major premise to generate a conclusion Q encoded and stored in working memory before being compared with the conclusions presented. When the conditional is uncertain (Thematic conditional), this uncertainty about the sufficiency of P for Q (or Q for P) seems to be taken into account by the subjects at the major premise and is manifested by an less pronounced expectation of Q conclusion when the premise P has been presented

Copy number variants (CNVs) account for both variations among normal individuals and pathogenic variations. The introduction of DNA microarrays had a significant impact on the resolution of detectable CNVs and yielded a new perspective on the submicroscopic CNVs. Oligonucleotide microarrays, such as Affymetrix SNP arrays, have been commonly used for genome-wide CNV analysis. Despite the improvements in the technology, a major concern of using microarrays is how a putative CNV is defined. A disadvantage of oligonucleotide arrays is the poor signal-to-noise ratio of the data that leads to considerable variation in reported intensity readouts. Such variation will lead to false positive and false negative results, regardless of how the data are analysed. The most common approach to circumvent this problem is looking for abrupt ratio intensity shifts in several consecutive markers (e.g., SNP probes). However this approach reduces the overall resolution and mitigates the sensitivity of detecting CNVs with fewer probes. This limitation emphasizes the importance of designing methods that can identify noisy readouts at the probe-level. The main goals of this work were to study the scale of the variability in Affymetrix SNP arrays and to develop computational tools that can improve the resolution of CNV detection. By using simulated data, it was shown that the proposed method improved the accuracy and precision of detecting CNVs with fewer probes compared to standard methods. This approach was also applied to tumor/normal pairs from 25 follicular lymphoma patients and 286 candidate CNVs were found, from which 261 (91.2%) were also seen by other array-based method(s). Importantly, from 32 novel deletions, undetected by other array-based methods, at least 15 (47%) were real based on sequence-based validation. An example of a novel discovery was a partial deletion of the extracellular domain of the KIT proto-oncogene that may lead to constitutive activation of this gene. Gain of function mutations of KIT has been previously reported in several other hematologic cancers through other mechanisms such as point mutations. In conclusion, CNV discovery contributes to our understanding of complex diseases and the methods presented here should provide means for better detection of CNVs and their interpretation.

Williams-Beuren Syndrome (WBS) and 7q11.23 microduplication syndrome (7dup), two multi-systemic developmental disorders, arise from symmetrical copy number variations of the same region on chromosome 7q. In WBS patients this region is deleted, whereas it is duplicated in 7dup individuals. These syndromes display a striking combination of shared and opposite clinical manifestations at the level of neuro-cognitive, craniofacial and cardiovascular features, thus pointing to a remarkable dosage-sensitive effect of a small group of genes on the development and maintenance of complex traits such as sociality, language and facial morphology. We derived a large cohort of induced pluripotent stem cells (iPSCs) from samples with WBS, 7dup and from healthy controls, and we demonstrated that, already at the pluripotent state, the transcriptome is dysregulated in pathways that map onto disease-related features. Moreover, these pathways were selectively dysregulated in differentiated lineages, thus demonstrating an anticipatory power of the pluripotent state. Building on these results, we expanded the view on the dysregulation in pluripotency by measuring three layers of gene expression: transcriptome, translatome and proteome. We mapped the propagation of differences across layers by integrating ribosome profiling and SWATH-MS proteomics, and we probed the extent to which a translation initiation factor included in the CNV, EIF4H, was responsible for the regulation of translation. We found that each layer of gene expression has its own differentially expressed genes, whose degrees of propagation can change between layers. Moreover, differentially expressed genes can cluster by different ways of propagation when they are compared to the levels of EIF4H.

Les variations en nombre de copies (CNV) sont des gains ou pertes d’une séquence d’ADN et peuvent avoir un rôle dans la susceptibilité à certaines maladies. Les CNVs peuvent être détectés par les puces de SNPs de haute résolution en analysant les intensités des allèles avec des algorithmes de détection des CNVs tels que CNV partition, PennCNV et QuantiSNP. Dans cette thèse, nous avons évalué les performances de ces outils pour la détection des CNVs au niveau pangénomique et pour les tests d'association. Nous avons également étudié des stratégies d'association combinant les informations de l'allèle et du nombre de copies pour des SNP situés dans des CNV. Nous avons appliqué ces outils pour mener une étude d’association pan-génomique avec les CNV en utilisant les données de l'étude espagnole du cancer de lavessie (SBC)/EPICURO générées par la puce Illumina 1M.Nos résultats montrent une faible fiabilité et une faible sensibilité des algorithmes de détection des CNV. Dans la région du gène GSTM1 où un CNV très fréquent existe qui est associé au risque de cancer de la vessie, nous avons constaté que les algorithmes de détection des CNV ont de faibles performances. Néanmoins, l’utilisation de la mesure d'intensité des allèles dans les tests d'association peut alors être une alternative intéressante car cela nous a permis de détecter cette association connue. Pour les SNPs situés dans des CNVs, nous avons étudié plusieurs stratégies de tests d'association et nous avons montré que la plus puissante était d’utiliser un modèle avec deux termes correspondant respectivement à la somme et à la différence du nombre de copies des deux allèles. Finalement, en appliquant ces stratégies à l'étude (SBC)/EPICURO, nous avons identifié des CNVs potentiellement associés au risque de cancer de la vessie, ainsi que des SNP dont l'allèle et le nombre de copies pourraient être impliqués dans le risque de cancer de la vessie
Copy number variations (CNVs) are losses or gains of DNA sequences that may play a role in specific disease susceptibility. CNVs can be detected by high-resolution SNP-arrays through the analysis of allele intensities with CNV calling algorithms such as CNVpartition, PennCNV and QuantiSNP. In this thesis, we identified and assessed the performances of available tools for CNV calling and for association testing, at the genome-wide level. We also investigatedassociation strategies that combine information on both the allele and the number of copies for SNPs located in CNV regions. We applied these tools to conduct a genome-wide association study with CNV using data from the Spanish Bladder Cancer (SBC)/EPICURO Study generated by the Illumina 1M SNP-array. Our results showed a low reliability and a low sensitivity of the investigated CNV calling algorithms applied to SNP-array data. The GSTM1 locus shows a very frequent CNV that is associated with bladder cancer (BC) risk. We reported that the calling algorithms performed very poorly in identifying this CNV. We proposed using allele intensity measures (LRR) as a screening step to assess association as it allowed the detection of the GSTM1 CNV association with BC. To combine the allele and the number of copies for SNPs located in CNV regions, we investigated several strategies of association testing and we showed that the more powerfulone used a two-term model with the sum and the difference of the number of copies of both alleles. Finally, by applying these strategies to the (SBC)/EPICURO Study, we identified CNV regions potentially associated with BC risk, as well as SNPs for which both the allele and the number of copies could be involved in BC risk

A complexidade da fisiologia da audição resulta da participação e interação de produtos de grande número de genes, razão pela qual a surdez hereditária exibe enorme heterogeneidade genética. Estudos moleculares nas duas últimas décadas permitiram a identificação de vários genes responsáveis por surdez; entretanto, muitos ainda restam ser identificados. A maioria dos estudos de mapeamento de genes de surdez até então conduzidos privilegiou estratégias que buscavam mutações de ponto. Outros mecanismos mutacionais, como deleções e duplicações, foram pouco investigados. Portanto, a contribuição das CNVs (Copy Number Variations) na surdez hereditária é pouco conhecida. O objetivo desse trabalho foi identificar novos genes que possam ter papel na etiologia da surdez sindrômica ou não-sindrômica por meio da investigação de microdeleções e microduplicações em pacientes com perda auditiva. Selecionamos 25 genes candidatos (CTTN, FGF3, FGF19, FOXC1, FOXF2, FOXQ1, IMMP2L, KIF5C, LRRN3, MAP1A, MYLK4, PPP3CA, SHANK2, SLC5A7, STRC, TMC1, TMC2, TMC3, TMC4, TMC5, TMC6, TMC7, TMC8, TPCN2 e TUBB2A) para a triagem de microrrearranjos por meio da técnica de MLPA (Multiplex Ligation-dependent Probe Amplification). Os genes candidatos foram selecionados a partir de rearranjos detectados em um estudo prévio realizado por meio de array-CGH (array-based Comparative Genomic Hybridization) em indivíduos com surdez sindrômica estudados em nosso laboratório, e também a partir de dados da literatura. Nossa casuística foi composta por 163 indivíduos, dos quais 74 são pacientes com surdez associada a outros sinais (sindrômicos), a maioria casos isolados, e 89 são pacientes com surdez não-sindrômica, propósitos de famílias em que segrega surdez de herança autossômica dominante ou recessiva. Desenhamos uma sonda sintética intragênica de MLPA para cada um dos genes candidatos. Foram detectadas seis deleções em TMC6 (3,7%), seis deleções e uma duplicação em STRC (4,3%) e uma duplicação em IMMP2L (0,6%). A triagem de alterações nesses três genes em 189 indivíduos fenotipicamente normais revelou quatro deleções em TMC6 (2,1%), oito deleções e três duplicações em STRC (5,8%) e três deleções em IMMP2L (1,6%). Todas as alterações em TMC6, tanto nos casos de surdez como nos controles, eram na realidade artefatos devidos a problemas de hibridação da sonda correspondente. No gene STRC, previamente já relacionado à surdez, os rearranjos nos indivíduos afetados devem se tratar de polimorfismos sem efeito fenotípico por serem muito frequentes na população. Contudo, é possível que haja nesses pacientes mutações adicionais que não puderam ser rastreadas e que poderiam contribuir ao fenótipo, em combinação com o rearranjo detectado, como já descrito em um caso da literatura. A duplicação em IMMP2L em uma paciente com surdez não-sindrômica, herdada da mãe igualmente afetada, mostrou-se a mais provavelmente relacionada ao fenótipo, pois o estudo complementar por meio de array-CGH revelou que o rearranjo inclui uma duplicação parcial da porção 3 de outro gene, DOCK4. O produto desse gene possui uma isoforma que se localiza nos estereocílios das células ciliadas e se liga a uma importante proteína relacionada à audição, a harmonina. Portanto, nossa hipótese é a de que a duplicação seja a causa da surdez na família e que DOCK4 seja um novo gene responsável por surdez. A associação de IMMP2L com surdez é menos provável devido ao grande número de CNVs não patogênicas já descritas que incluem partes desse gene. Estudos complementares são necessários para mapear a duplicação com mais precisão. Além disso, o rastreamento de mutações em DOCK4 em outras famílias com surdez pode vir a confirmar o possível papel desse gene na etiologia da surdez.
Several genes contribute to the complexity of physiology of hearing. Consequently, hereditary deafness is extremely heterogeneous from the genetic point of view. In the last two decades, several genes responsible for hereditary hearing loss have been identified, but a large number of genes remains to be found, as evidenced by the unexplained cases of inherited deafness. The search for point mutations in candidate genes after mapping based on linkage studies has been the main strategy in the identification of such genes. Other mutation mechanisms, such as deletions and duplications, have been rarely investigated, and the contribution of DNA copy number variants (CNVs) to hearing loss is not well known. This study aimed at identifying novel genes, which might play a role in the etiology of syndromic and non-syndromic deafness, through the search of gene microdeletions and microduplications. We selected 25 candidate genes (CTTN, FGF3, FGF19, FOXC1, FOXF2, FOXQ1, IMMP2L, KIF5C, LRRN3, MAP1A, MYLK4, PPP3CA, SHANK2, SLC5A7, STRC, TMC1, TMC2, TMC3, TMC4, TMC5, TMC6, TMC7, TMC8, TPCN2 and TUBB2A) based on their involvement in microimbalances detected by Array-based Comparative Genomic Hybridization (aCGH) in a previous study of a Brazilian sample of individuals with syndromic hearing loss from our laboratory and others reported in the literature. We studied 163 subjects, 74 of them presenting syndromic deafness, the majority were isolated cases, and 89 being probands of families in which nonsyndromic deafness had an autosomal dominant or recessive mode of inheritance. Gene deletions or duplications were screened by Multiplex Ligant-dependent Probe Amplification (MLPA) using one synthetic intragenic probe designed for each candidate gene. We detected six deletions in TMC6 (3,7%), six deletions and one duplication in STRC (4,3%), and one duplication in IMMP2L (0,6%). The screening of imbalances in these genes in a control sample of 189 hearing individuals revealed four deletions in TMC6 (2,1%), eight deletions and three duplications in STRC (5,8%) and three deletions in IMMP2L (1,6%). The imbalances found in TMC6, both in affected and control individuals, were in fact artifacts due to problems in the hybridization of the corresponding probe. As to the STRC gene, previously related to deafness, the imbalances are more likely to be 4 polymorphisms with no phenotypic effect. However, the possibility remains that additional undetected mutations in affected individuals contribute to their phenotype, in combination with the microrearrangement, as already reported in the literature. The duplication in IMMP2L in a non-syndromic patient, and also present in her affected mother, is most likely causative of deafness, since a complementary study performed with aCGH revealed that the rearrangement included a partial duplication of the 3 end of another gene, DOCK4. An isoform of the DOCK4 protein localizes to the stereocilia in the inner ear and interacts with harmonin, a protein already known to be involved in hearing. We hypothesize that this duplication may be the cause of deafness in the family and, this being the case, DOCK4 appears as a novel deafness gene. The causal association between IMMP2L and deafness is less plausible, because of the large number of reported non-pathogenic CNVs that include parts of this gene. Further studies are required to precisely map this duplication. In addition, the screening of mutations in DOCK4 in other families with hearing impairment is required to evaluate its possible role in the etiology of deafness.

Le FcRn est le récepteur responsable du recyclage des IgG ainsi que de leur transcytose. Ainsi cette protéine a un rôle majeur dans la pharmacocinétique des anticorps thérapeutiques. Nous nous sommes intéressés à différents aspects de l'expression du FcRn. Premièrement nous avons évalué l'influence sur la pharmacocinétique du cétuximab de 2 polymorphismes génétiques influençant l'expression du FcRn. Nous avons montré qu'un Variable Number Tandem Repeat influence la distribution du cétuximab. Nous avons établi que le gène est rarement soumis à des variations de son nombre de copies. Par ailleurs, nous avons montré par une approche de RT-PCR en multiplex l'absence du transcrit FcRn dans les plaquettes humaines. Enfin, l'analyse du niveau de transcrits de FcRn dans un modèle d'activation cellulaire indique qu'il existe une régulation: ce niveau diminue lorsque des monocytes sont différenciés en cellules dendritiques immatures ainsi que lors de l'évolution en cellules dendritiques matures. Les résultats de cette thèse démontrent l'importance de l'étude de l'expression du FcRn dans la variabilité pharmacocinétique des anticorps thérapeutiques
The FcRn is the receptor responsible for the recycling of IgG and their transcytosis. Thus, this protein has a major role in the pharmacokinetics of therapeutic antibodies. We focused on different aspects of FcRn expression. First, we evaluated the influence on the pharmacokinetics of cetuximab of 2 genetic polymorphisms influencing FcRn expression. We showed that a Variable Number Tandem Repeat influences the distribution of the cetuximab. We determined that the gene is rarely affected by Copy Number Variations. Furthermore, we showed by an RT-PCR approach that the FcRn transcript is absent in human platelets. Finally, the analysis of FcRn transcript level in a model of cellular activation indicates that a regulation occurs : the level decreases when monocytes differenciate into immature dendritic cells as well as during evolution into mature dendritic cells. Results of this thesis demonstrate the importance of the study of FcRn expression in pharmokinetic variability of therapeutic antibodies

A infertilidade é um problema de saúde pública com um significativo impacto social, econômico e psicológico. Em todo o mundo, a incidência da infertilidade entre a população geral é estimada em 10-15%. Cerca de 50% da infertilidade dos casais são de origem masculina. Em mais da metade dos homens inférteis, a causa da infertilidade é desconhecida (idiopática). Etiologicamente, a infertilidade masculina apresenta causas genéticas e não genéticas. Dentre as causas genéticas mais conhecidas temos mutação do receptor de andrógenos, mutação do gene regulador da condutibilidade transmembrana da fibrose cística (CFTR), anomalias cromossômicas clássicas, anomalias meióticas, microdeleções do cromossomo Y, etc. As anomalias cromossômicas são encontradas com muito mais frequência em homens inférteis, com uma incidência de 4-16% em relação à incidência de 0,4% na população fértil. Estudos mostram que as CNVs também podem estar relacionadas com a infertilidade masculina, especificamente com a falha na espermatogênese. CNVs encontradas tanto no cromossomo Y quanto nos cromossomos autossômicos também foram associadas a possíveis falhas na espermatogênese. Um outro fator que também pode estar envolvido com a infertilidade masculina é a expressão desregulada dos miRNAs. O presente trabalho teve como objetivo promover a análise em larga escala da distribuição de CNVs e do perfil transcricional dos miRNAs em amostras de biopsias testiculares de paciente com azoospermia. Para o estudo das CNVs nós utilizamos a metodologia do CytoScan HDTM da Affymetrix. O perfil transcricional de miRNAs nos indivíduos estudados foi avaliado por meio da tecnologia de microarranjos também da plataforma Affymetrix. Para estas analises montamos dois grupos de estudo (Parada de Maturação (MA) de Células Germinativas e Síndrome de Células Sertoli Only (SCOS)) e um grupo controle (azoospermia obstrutiva e espermatogênese normal). Através das análises das CNVs nós encontramos 94 CNVs nos cromossomos autossômicos e sexuais, 35 (37%) CNVs foram classificadas como benignas, 24 (23%) como potencialmente benignas, sete CNVs (7,4%) como patogênicas e sete foram classificadas como potencialmente patogênica. Todas as CNVs classificadas como patogênica estão presentes no cromossomo Y, cinco CNVs são do tipo duplicação e duas do tipo deleção. A CNV do tipo duplicação foi encontrada no paciente MA e a CNV do tipo deleção foi encontrada no paciente SCOS. As CNVs se sobrepõem e quando analisadas em conjunto (formando uma única CNV de cada condição) elas apresentam um tamanho parecido. Estas CNVs apresentam genes envolvidos na espermatogênese. As CNVs classificadas como potencialmente patogênicas estavam presentes nos cromossomos autossômicos e cromossomo X. Nestas CNVs estavam presentes genes que foram associados com a falha na espermatogênese. A análise da expressão dos miRNAs revelou um perfil transicional muito mais alterado nos pacientes com SCOS. As duas condições apresentaram miRNAs exclusivos, mas também compartilharam: 30 miRNAs. Nós identificamos duas famílias de miRNAs (miR449 e miR34) diferencialmente expressos nas duas condições e que apresentam expressão preferencial no testículo. Nossos resultados mostram que alterações no número de copias (CNVs) no cromossomo Y levam a infertilidade masculina e CNVs nos cromossomos autossômicos e X podem levar a infertilidade masculina. As alterações do tipo deleção podem levar a uma falha na espermatogênese maior que as alterações do tipo duplicação. A expressão diferencial dos miRNAs em tecido testicular de pacientes com diferenças histopatológicas (SCOS e MA) apresentam um padrão de expressão de miRNAs diferentes devido ao tipo de células germinativas que eles apresentam no tecido epitelial do testículo.
Infertility is a public health problem with significant social, economic and psychological impact. Worldwide, the incidence of infertility in the general population is estimated at 10- 15%. Approximately 50% of infertility of couples is of male origin. In more than half of infertile men, the cause of infertility is unknown (idiopathic). Etiologically, male infertility has genetic and non-genetic causes. Among the best known genetic causes we found the mutation of the androgen receptor, the cystic fibrosis transmembrane conductance regulator (CFTR), classic chromosomal abnormalities, meiotic abnormalities and microdeletions of the Y chromosome. Chromosomal abnormalities are found much more frequently in infertile men, with an incidence of 4-16% in the incidence of 0.4% in the fertile population. Studies show that CNVs can also be related to male infertility, specifically in the failure of spermatogenesis. CNVs found in both the Y and autosomes chromosomes were also associated with possible failures in spermatogenesis. Another factor that may also be involved in male infertility is the deregulated expression of miRNAs. This work aimed to promote the analysis of large-scale distribution of CNVs and the transcriptional profile of miRNAs in testicular biopsy samples from patients with azoospermia. For the study of CNV we used the CytoScan HDTM Affymetrix methodology and the transcriptional profile of miRNAs in the samples was assessed by means of microarray technology from Affymetrix platform. For these analyzes we set up two study groups (Stop Maturation (MA) of Germ Cells and Sertoli Cell Only Syndrome (SCOS)) and compared them to a control group (obstructive azoospermia, normal spermatogenesis). Through analysis of CNVs, we found 94 CNVs in sexual and autosomes chromosomes, 35 (37%) were classified as benign CNVs, 24 (23%) as a potentially benign seven CNVs (7.4%) as pathogenic and 7 were classified as potentially pathogenic. All CNVs classified as pathogenic are present on the Y chromosome, five CNVs are of duplication type and two are deletion type. The duplication type CNV was found in MA patients and deletion type CNV was found in SCOS patient. We identified that CNVs overlap and when analyzed jointed - as a single CNV of each condition - they have a similar size. These CNVs have genes involved in spermatogenesis. CNVs classified as potentially pathogenic were present in autosomes and in the X chromosome. In these CNVs were present genes that were associated with failure in spermatogenesis. The analysis of the expression of miRNAs revealed a transitional profile much more altered in patients with SCOS. The two conditions presented exclusive miRNAs, but shared 30 miRNAs differentially expressed when compared to the control group. We identify two families of miRNAs (miR449 and miR34) which exhibit preferential expression in testis as differentially expressed in both conditions. Our results show that changes in the number of copies (CNVs) on the Y chromosome lead to male infertility and CNVs in autosomes and X chromosomes may lead to male infertility. The deletion type changes can lead to a failure of spermatogenesis greater than the duplication type changes. The differential expression of miRNAs in patients with testicular tissue histopathologic differences (SCOS and MA) has a different pattern of miRNA expression due to the type of germ cells they present in epithelial tissue of the testis.

Le macrosatellite RNU2 est constitué de répétitions en tandem d'une unité de 6,1 kb. Largement étudié pendant les années 1980 et 1990, il est maintenant oublié des études pan-génomiques du fait de son absence du génome de référence. J'ai dans un premier temps finement caractérisé ce macrosatellite, en réalisant un assemblage in silico de la région génomique, en développant un code-barres pour la technique de peignage moléculaire et en analysant les données du projet 1000 Génomes. J'ai ainsi validé la localisation du locus RNU2 124 kb en amont de BRCA1, et affiné les données de polymorphisme en montrant que le nombre allélique de copies pouvait varier entre 5 et 82 chez 42 individus. J'ai tiré profit de sa localisation au sein d'un large bloc de déséquilibre de liaison pour définir le taux de mutation de ce macrosatellite à l'origine du nombre important d'allèles identifiables au sein de la population générale. Compte tenu de sa proximité avec BRCA1 et de son fort taux de polymorphisme, j'ai étudié le nombre global de copies du CNV dans 2 cohortes de cas de cancer du sein et témoins associés. J'ai montré que le nombre global de copies est significativement plus élevé chez les cas que chez les témoins. Ce travail suggère que le nombre de copies du macrosatellite RNU2 pourrait être impliqué dans la prédisposition génétique au cancer du sein, impliquant ainsi pour la première fois un CNV dans un mécanisme d'inactivation d'un gène de prédisposition au cancer

A perda auditiva é o defeito mais comum ao nascimento e cerca de 70 milhões de pessoas no mundo apresentam algum grau de perda auditiva. Além da alta incidência, as implicações da perda auditiva na linguagem, na cognição e no desenvolvimento emocional e social reforçam sua importância. No entanto, em grande parte dos pacientes, a causa da deficiência auditiva não é esclarecida. Nós usamos hibridação comparativa do genoma baseada em arrays (Array Comparative Genomic Hybridization aCGH) para investigar alterações no número de cópias de segmentos de DNA (Copy Number Variation CNV) em 31 indivíduos que apresentavam deficiência auditiva e sinais clínicos adicionais, mas que não puderam ser classificados em síndrome conhecida. A escolha de indivíduos sindrômicos se baseou no pressuposto de que, em média, apresentam alterações genômicas maiores e, portanto, mais provavelmente detectáveis com o uso de aCGH de 1 Mb, que era a plataforma disponível no início do projeto. CNVs não descrita em bancos de dados de indivíduos normais foram identificadas em oito pacientes, quatro delas ocorreram de novo enquanto as outras quatro foram herdadas de um genitor fenotipicamente normal. As alterações de novo definem segmentos cromossômicos que provavelmente contém genes relacionados à deficiência auditiva e sensíveis a dose, especificamente: 1q23.3-q25.2, 2q22q23, 6p25.3 e 11q13.2-q13.4. As alterações raras identificadas tanto nos pacientes quanto em um genitor normal poderiam ser um evento ao acaso, sem papel na deficiência auditiva; no entanto, a possibilidade de que essas alterações possam funcionar como fatores de predisposição não podem ser descartadas. Se considerarmos apenas as CNVs de novo como causativas dos fenótipos investigados, detectamos quatro pacientes portadores entre os 31 investigados (13%). Se considerarmos também as CNVs herdadas como possivelmente causativas, a taxa de desequilíbrios cromossômicos associados à surdez será de 26%. Esses resultados são provavelmente uma substimativa e esses números seriam possivelmente maiores com o uso de uma das plataformas de alta resolução disponíveis atualmente. Esses resultados, embora limitados, indicam que investigação por aCGH em pacientes com surdez sindrômica idiopática está entre os testes mais eficientes para detectar etiologia dos fenótipos, devendo ser incorporado à rotina no diagnóstico e aconselhamento genético.
Hearing loss is the most common congenital deficiency and about 70 million people worldwide present some degree of hearing impairment. In addition to its high incidence, hearing loss impacts language, cognition and social and emotional development. However, in a large proportion of patients, the cause of the hearing deficiency cannot be elucidated. We screened copy number changes by 1 Mb-array Comparative Genomic Hybridization (aCGH) in 31 individuals with syndromic hearing impairment whose clinical features were untypical for known disorders. The choice of evaluating syndromic rather than non-syndromic individuals was based on the assumption that they are more likely to carry larger genomic alterations which could be more easily detected by the comparatively low resolution 1 Mb aCCG, which was the available platform when this project started. Copy number changes (CNV) not documented in the database of normal individuals were detected in eight patients, four de novo imbalances and four inherited from a normal parent. The de novo alterations define candidate chromosome segments likely to harbor dosage sensitive genes related to hearing impairment, namely 1q23.3-q25.2, 2q22q23, 6p25.3 and 11q13.2- q13.4. The rare imbalances also present in normal parents might be casually associated with hearing impairment, but also have a possible role as a predisposition factor. When only the de novo CNVs were considered causative for the disease phenotypes, our study revealed relevant copy number changes in 4 patients (13%). If we also count the rare CNVs that had been inherited as possibly causative, the frequency of chromosome imbalances associated with syndromic deafness in our sample becomes 26%. These figures are probably underestimates and will probably become larger when high resolution oligoarray platforms are applied. These results indicate that aCGH is an efficient tool for defining the etiology of syndromic deafness and its use in routine diagnosis of hearing impairment and for genetic counseling is highly recommended.

La vigne (Vitis vinifera) possède un métabolisme secondaire particulièrement riche donnant naissance à une large palette de molécules dont certaines sont impliquées dans les défenses contre les pathogènes et d'autres dans la grande diversité d’arômes qui fait la renommée des vins. L’analyse de la séquence de référence du génome de la vigne a permis de mettre en évidence une remarquable expansion de certaines familles de gènes liés au métabolisme secondaire par rapport aux autres plantes. Dans ce travail, j'ai étudié les familles gènes codant pour les cytochromes P450, dont certains sont impliqués dans la production d’arômes, les gènes codant pour les stilbènes synthases (STS), les endo-β-1,3-glucanases et les gènes de résistance de type NBS impliqués dans les défenses de la vigne. Ma thèse vise à proposer des hypothèses expliquant l’organisation structurale de ces familles de gènes et ainsi à mieux comprendre pourquoi certaines familles présentent une amplification dans le génome de la vigne. Des approches bioinformatiques ont été utilisées afin d’étudier ces différentes familles de gènes. Les gènes cytochromes P450 et gènes R de type NBS ont tout d'abord été annotés de manière manuelle dans le génome de référence de la vigne. L’expression des gènes endo-β-1,3- glucanases, STS et cytochromes P450 a été analysée en utilisant une approche transcriptomique à grande échelle. Pour ce faire, un outil a été développé durant cette thèse pour estimer le niveau d’expression des gènes à partir de données RNA-Seq disponibles dans les banques de données publiques. Parallèlement, des données de reséquençage d’ADN de 56 cépages et espèces de vigne ont été analysées, afin de déterminer les variations structurales de type CNV au sein des familles de gènes à domaine NBS et de gènes STS. Ces différents travaux ont permis de montrer que l’amplification des familles de gènes étudiées n’est pas spécifique du génome de référence mais est retrouvée dans l'ensemble du genre Vitis, mais également de mettre en évidence des variations structurales au sein des différents génomes étudiés. L'analyse de la famille STS a montré que ces gènes sont organisés en blocs de duplication, et que les gènes plus conservés sont aussi les plus exprimés. Nous avons également montré que les gènes à domaine NBS sont organisés en cluster, dont certains sont particulièrement soumis à variation. Ces travaux contribuent à une meilleure connaissance de facteurs de défense efficaces et durables ainsi que des gènes impliqués dans la synthèse d’arômes dans la vigne. Ces connaissances pourront bénéficier aux programmes de création variétale mis en œuvre à l’INRA de Colmar
Grapevine (Vitis vinifera) has a particularly rich secondary metabolism, giving rise to a wide range of molecules, some of which are involved in defences against pathogens and others in the great diversity of aromas that make wines famous. Analysis of grapevine reference genome has shown a remarkable expansion of certain families of genes linked to secondary metabolism in comparison with the other plants. In this work, I have analysed gene families coding for cytochromes P450, some of them being involved in the production of aromas, genes coding for stilbene synthases (STS), endo-β-1,3-glucanases and NBS type resistance genes involved in grapevine defences. My thesis intends to propose hypothesis to explain the structural organisation of these families and therefore better understand why some of these families are amplified in the grapevine genome. Bioinformatic approaches have been used to study these different genes families. The cytochromes P450 and R genes of NBS type were manually annotated to improve the knowledge of these families of genes. The expression of endo-β-1,3-glucanases, STS and cytochromes P450 genes has been quantified using a large-scale transcriptomic approach. To this purpose, a tool has been developed during this thesis to estimate the level of genes expression from RNA- Seq data available in public databases. In the meantime, DNA resequencing data from 56 cultivars and grapevine species have been analysed to identify structural variations of CNV types within the genes with a NBS domain and the STS genes. These works showed that the amplification of the gene families of interest was not specific to the reference genome but occurred at the scale of the Vitis genus, but also to highlighted structural variations in different genomes. Regarding the STS genes, blocks of duplication and more conserved and expressed genes were identified. For the genes with NBS domain, a clustered organisation has been highlighted with some clusters varying more than others in the studied genotypes. These works contribute to a better knowledge of gene families for efficient and durable defence against pathogens and optimal aromas synthesis in grapevine. This knowledge will benefit to breeding programs currently in progress at INRA Colmar

Ziel der vorliegenden Arbeit war es, mittels des Aufgabenwechselparadigmas, kognitive Prozesse nicht nur anhand von traditionellen Leistungsparametern, sondern zusätzlich durch elektro-physiologische Parameter zu untersuchen. Parameter ereigniskorrelierter Hirnpotentiale (EKP) wurden ebenfalls zur Einschätzung von altersbedingten Änderungen bei der Ausführung von Reaktionszeitaufgaben herangezogen.
Nach Rubinstein et al. (2001) setzt sich die Reaktionszeit aus der Dauer seriell angeordneter Verarbeitungsstufen zusammen. Im Stufenmodell der exekutiven Kontrolle von Rubinstein et al. (2001) sind Prozesse der ausführenden Kontrolle nur an Wechseltrials beteiligt und können getrennt von den Aufgabenprozessen ablaufen. Mittels der Informationen zu den Reaktionszeiten ist es jedoch nicht möglich zu klären, auf welche kognitiven Verarbeitungsprozesse Reaktionszeitunterschiede unter den jeweiligen experimentellen Bedingungen zurückzuführen sind. Zur Analyse der kognitiven Prozesse wurden in dieser Untersuchung die CNV und P300 herangezogen. Es wurden zwei Altersgruppen (20-30 Jährige und 49-61 Jährige) untersucht. Den Probanden wurden Ziffern präsentiert, die entweder nach dem numerischen Wert oder der Schriftgröße mit dem Hinweisreiz, der Zahl 5, verglichen werden sollten. Die Stimuli wurden nach dem Alternating-Runs-Paradigma dargeboten (Rogers und Monsell, 1995).
Erwartungsgemäß gab es Reaktionszeitunterschiede zwischen alt und jung mit längeren Reaktionszeiten für die älteren Probanden. Altersunterschiede in den Fehlerraten ließen sich nicht nachweisen. Möglicherweise erfolgte die Reaktionsauswahl bei den Älteren überlegter aus als bei den Jüngeren. Dies spiegelte sich in längeren aber fehlerfreien Reaktionen wider. Vermutlich bereiteten jedoch alle Probanden in dem Intervall zwischen Cue und Stimulus das jeweilige Aufgabenset komplett vor. Das könnte auch erklären, warum es bei einem Aufgabenwechsel nicht zu einem Anstieg der Reaktionszeit und der Fehlerrate kam. Entgegen der Erwartung zeigten sich keine Wechselkosten. Teilweise wurden inverse Wechselkosten nachgewiesen. In Bezug auf die Wechselkosten konnte das Stufenmodell der exekutiven Kontrolle (Rubinstein et al., 2001) nicht bestätigt werden. Der explizite Hinweisreiz scheint allerdings Einfluss auf die Wechselkosten zu haben. Verschiedene Erklärungsansätze werden diskutiert.
Die Contingent Negative Variation ist wie erwartet vor einem Aufgabenwechsel größer als vor einer Aufgabenwiederholung. Durch den Hinweisreiz ist eine erhöhte Kapazität vorhanden. Entsprechend den Ergebnissen der CNV kann davon ausgegangen werden, dass ältere Erwachsene stärker von der Vorinformation zu profitieren scheinen als jüngere Erwachsene. Die älteren Erwachsenen beginnen im Gegensatz zu den jüngeren Erwachsenen offenbar eher mit der Vorbereitung. Zeitdruck und Aufgabenwechsel lösen eine stärkere P300 aus. Demzufolge scheinen Zeitdruck und Aufgabenwechsel einen erhöhten Kapazitätsbedarf zu erfordern. Im Sinne des Stufenmodells der exekutiven Kontrolle von Rubinstein et al. (2001) führt die Zielverschiebung bei einem Aufgabenwechsel zu einer größeren P300. Die Resultate der hier dargestellten Untersuchungen verdeutlichen, dass ältere Erwachsene einen höheren zeitlichen Aufwand in den Stufen der einzelnen exekutiven Prozesse benötigen. Dies spricht für die Hypothese der selektiven Verlangsamung. Ältere kompensieren dies durch einen höheren Aufwand in der Vorbereitung, was auf elektrokortikaler Ebene nachweisbar ist, sind aber nicht in der Lage, dies in den Reaktionszeiten umzusetzen.
Die Ergebnisse dieser Untersuchung unterstützen die vereinfachte Annahme von Rubinstein et al. (2001), nach dem die Teilprozesse der Reaktionszeit seriell verarbeitet werden können. Die Resultate lassen allerdings den Schluss zu, dass die Wechselkosten im Hinblick auf die Reaktionszeiten nicht der geeignete Parameter für die Messung der exekutiven Kontrolle sind.
Die vorgeschlagene Modifikation des Modells von Rubinstein et al. (2001) in der Vorbereitung auf eine Aufgabe gilt es in weiteren Untersuchungen zu bestätigen und die Möglichkeit der Anwendung auf alle Aspekte der exekutiven Kontrollprozesse zu prüfen.
The aim of this study was it to examine cognitive processes not only on the basis achievement parameters by means of the alternating runs paradigm, but additionally by electricalphysiological parameters. Parameters of event-correlated brain potentials (EKP) were also used to estimate age-related changes in tasks of response time.
According to Rubinstein et al. (2001) the response time consists of the duration of serially arranged processing levels. In Rubinstein's et al. (2001) stage model of the executive control processes of implementing control are involved only in switch trials and can run separately from the task processes. The information from response times do not aloud to define what cognitive processing processes are responsible for response time differences in respective to the experimental conditions. In this study the contingent negative variation (CNV) and P300 were used for the analysis of the cognitive processes.
Two age groups (20 to 30, and 49 to 61 years old) were included in the study. Numbers were presented, which should be compared to the cue number 5 either to the numeric value or character size. The stimuli were represented after the alternating runs paradigm (Rogers & Monsell, 1995).
As expected there were response time differences between old and young subjects with longer response times for the older ones. Age differences in the error rates could not be proven. It is possible that the reaction selection in older ones took place with more consideration than in the younger ones. This is reflected in longer but error free reactions. Probably all subjects prepared in the interval the respective task set between cue and stimulus completely. This could also explain, why with a task switching no rise of the response time and the error rate was noticed. Against expectation no switch costs showed up. Inverse switch costs were partly proven. Regarding the switch costs the stage model of the executive control (Rubinstein et al., 2001) could not be confirmed. The explicit cue however seems to have influence on the switch costs. Different explanations are discussed.
The contingent negative variation is higher before a task switching than before a repetition of task. By the cue an increased capacity is presented. According to the results of the CNV it can be assumed that older adults seem to profit more strongly than younger adults from the advance information. Obviously the older adults begin earlier with the preparation than the younger ones. Time pressure and task switching release a stronger P300. Therefore time pressure and task switching seem to require an increased capacity need. According to the stage model of the executive control (Rubinstein et al., (2001) the goal shift with goes along with the task switching leads to a higher P300.
The results of the study represented here clarify, that older adults need a more time in the stages of the individual executives of processes. This underlines the hypothesis of the selective slowing down. Older ones compensate this by more effort to preparation, which can be proven on the electrocortical level. The are not able to show this in the the response times, though. The results of this study support the simplified acceptance of Rubinstein et al. (2001), after which the subprocesses of the response time can be processed serially.
The results permit the conclusion that the switch costs regarding the response times are not the adaquate parameter for the measurement of the executive control.
The suggested modification of the model of Rubinstein et al. (2001) in the preparation for a task it applies to be confirmed in further investigations and the possibility of application to all aspects of the executives has to be tested.

A displasia frontonasal (DFN) compreende quadros de aparência facial variável, sendo clinicamente caracterizada por dois ou mais dos seguintes sinais: hipertelorismo ocular com consequente alargamento da base nasal; fissura facial mediana afetando o nariz ou o nariz e lábio superior e, por vezes, o palato; fissura alar (uni ou bilateral); ponta nasal ausente; crânio anterior bífido oculto, e implantação em 'V' dos cabelos na fronte. A DFN pode ser vista como um defeito de desenvolvimento que pode ocorrer por si só ou como parte do quadro clínico de várias síndromes. A maioria dos casos de DFN é esporádica, e em raras circunstâncias foram observadas alterações cromossômicas em alguns indivíduos. Até o momento, quatro genes foram relacionados à patogênese molecular de algumas das síndromes com DFN, EFNB1, associado a uma forma de DFN ligada ao X e os genes ALX1, ALX3 e ALX4, todos associados a formas de DFN com herança autossômica recessiva. Embora esteja claro haver heterogeneidade etiológica, na maioria dos casos de DFN a causa não é conhecida, dificultando o adequado aconselhamento genético aos pacientes e seus familiares. Sendo assim, realizamos estudos com diferentes estratégias metodológicas buscando melhor compreender as possíveis causas genéticas da DFN. Ao todo foram analisados 10 pacientes: um caso familial de DFN leve com herança aparentemente autossômica dominante, um caso clinicamente sugestivo de mutação em ALX1, e oito casos de DFN associada a atraso de desenvolvimento com ou sem outras anomalias, dos quais um apresentava um rearranjo de novo aparentemente balanceado entre os cromossomos 4 e 12. Optamos por realizar sequenciamento dos genes previamente relacionados a fenótipos com DFN em todos os casos; para aqueles em que não foram detectadas mutações patogênicas, realizamos análise de variações de número de cópias (CNV) por microarray de polimorfismos de base única e, para o paciente com rearranjo cromossômico, realizamos o mapeamento do ponto de quebra por hibridação in situ fluorescente. Constatamos uma mutação em heterozigose no gene ALX4 co-segregando com o fenótipo do caso familial, sendo esta a primeira descrição de alteração em tal gene causando uma forma de DFN com herança dominante, e sugerimos pela primeira vez um mecanismo de dominância negativa. No caso sugestivo de mutação em ALX1, o diagnóstico foi confirmado através da identificação de uma mutação em homozigose neste gene do paciente; este caso consiste no 3o da literatura mundial e evidencia pela primeira vez que mutações em ALX1 não necessariamente levam a atraso de desenvolvimento ou deficiência intelectual. Os estudos citogenéticos e moleculares dos pontos de quebra do paciente com rearranjo cromossômico sugeriram os genes ARAP2 e CAND1 como possíveis responsáveis por seu quadro clínico, enquanto o estudo de CNVs nos indivíduos com DFN associada a atraso de desenvolvimento apontou os genes DNAJB12 e ENOX2 como possíveis candidatos para explicar o fenótipo de dois dos pacientes. É preciso que novos estudos sejam realizados a fim de melhor compreender o significado de tais achados e a real contribuição de cada gene para o desenvolvimento craniofacial humano e para a etiologia da DFN. Para os casos em que não foram identificadas alterações conclusivas no presente estudo, embora causas ambientais não possam ser descartadas, é preciso que seja investigada também a existência de fatores genéticos e epigenéticos não detectáveis pelas metodologias utilizadas, bem como a hipótese de mosaicismo somático. Nossos resultados, além de corroborarem o envolvimento dos genes ALX1 e ALX4 em fenótipos com DFN, sugerem também novos genes candidatos: ARAP2, CAND1, DNAJB12 e ENOX2
Frontonasal dysplasia (FND) is a rare group of disorders that comprises cases with a variety of facial appearances, and is clinically characterized by two or more of the following signs: ocular hypertelorism with consequent broadening of the nasal root; median facial cleft affecting the nose and/or upper lip and palate; clefting of the alae nasi (uni or bilateral); lack of formation of the nasal tip; anterior cranium bifidum occultum; and a V-shaped frontal hairline. FND is a developmental defect that can occur alone or as part of several syndromes. Most cases of FND are sporadic, and in rare circumstances chromosomal alterations were observed in affected individuals. To date, four genes have been related to the molecular pathogenesis of some syndromes with DFN, one (EFNB1) is associated with an X-linked form while the 3 others (ALX1, ALX3 and ALX4) are associated with autosomal recessive forms. Although it is clear that FND is etiologic heterogeneous, the causative mechanism is unknown in most cases which makes it hard to give proper genetic counseling to patients and their families. In order to get new insights into the genetic mechanisms leading to FND, we performed studies with different methodologies. Altogether, 10 patients were analyzed: a familial case of a mild form of FND with an apparently autosomal dominant inheritance pattern, a case clinically suggestive of mutation in ALX1, and eight cases of FND associated with developmental delay with or without other anomalies, one of which with an apparently balanced de novo rearrangement between chromosomes 4 and 12. We chose to sequence the genes previously associated with FND phenotypes in all cases; for those in which pathogenic mutations were not detected, we conducted an analysis of copy number variations (CNV) by single nucleotide polymorphisms microarrays; for the patient with chromosomal rearrangement, we also mapped the breakpoints by using fluorescence in situ hybridization. We found a heterozygous mutation in ALX4 co-segregating with the phenotype of the familial case; this is the first description of mutation in this gene causing a form of FND with dominant inheritance pattern, and we suggested for the first time a dominant negative mechanism. In the case suggestive of mutation in ALX1, the diagnosis was confirmed by the identification of a homozygous mutation in this gene; this is the third case of the literature and shows for the first time that mutations in ALX1 are not necessarily related to developmental delay or intellectual disability. Breakpoints cytogenetic and molecular studies done with the patient with chromosomal rearrangement suggested ARAP2 and CAND1 genes as causative candidates for his condition, while the study of CNVs in individuals with FND associated with developmental delay pointed DNAJB12 and ENOX2 genes as possible candidates to explain the phenotypes of two of the patients. Further studies are necessary to better understand the significance of such findings and the actual contribution of each of these genes to human craniofacial development and the etiology of FND. Although environmental causes cannot be ruled out, it should also be investigated the existence of genetic and epigenetic factors as well as the possibility of somatic mosaicism, among the cases negative for the molecular approaches used in our study. Our results corroborate the involvement of ALX1 and ALX4 in FND phenotypes, and suggest new candidate genes: ARAP2, CAND1, DNAJB12 and ENOX2.

MicroRNAs (miRNAs) are studied as key genetic elements that regulate the gene expression involved in different human diseases. Clinical sequence based variations like copy number variations (CNVs) affect miRNA biogenesis, dosage and target recognition that may represent potentially functional variants and relevant target bindings. To systematically analyze miRNA-related CNVs and their effects on related genes, a user-friendly free online database was developed to provide further analysis of co-localization of miRNA loci with human genome CNV regions. Further analysis pipelines such as miRNA-target to estimate the levels or locations of variations for genetic duplications, insertions or deletions were also offered. Such information could support the simulation of miRNA-target interactions.

My thesis is focused on sequencing-based methods for lung diseases monitoring. Modern sequencing techniques allow us to comprehensively characterize nucleic acids obtained from patient-derived biological material, with potential applications in both basic research and clinical practice. The thesis is divided in two sections: In Section 1: “Evidence for host-dependent RNA editing in the transcriptome of SARS-CoV-2” I describe the presence of RNA editing events in SARS-CoV-2, by analysing publicly available second generation RNA sequencing data from infected patients: Emerging viral infections represent a threat to global health, and the recent outbreak of novel coronavirus disease 2019 caused by SARS-CoV-2 exemplifies the risks. RNA editing is a physiological mechanism mediated by two enzyme families: APOBEC and ADAR, which introduce A-to-I and C-to-U mutations in double strand and single strand RNA respectively. RNA editing typically involves endogenous RNAs but, if targeting viral RNA, it is potentially deleterious for virus’ viability itself, by generating premature stop codons and missense mutations in the viral genome. On the other hand, RNA editing could fuel virus evolution by increasing the basal mutational rate. I have downloaded publicly available Illumina transcriptomic data, from BALF samples of infected patients; using a combination of published tools (Reditools2, JACUSA) I have detected an enrichment of APOBEC and ADAR related mutation on SARS-CoV-2 RNA. A similar enrichment was observed in genomic RNA SARS-CoV-2, SARS and MERS sequences downloaded from GISAID and NCBI virus. The evidence of RNA editing on SARS-CoV-2 suggests that APOBEC and ADAR can interact with viral RNAs, probably with an anti-viral purpose. C-to-U changes leading to stop codons are overrepresented in the transcriptomic data but—as expected—disappear in the genomic dataset. This might point—again—to an antiviral role for these editing enzymes. Also, the proportion of ADAR-related mutation was unexpectedly lower in genomic RNA sequences, compared to transcriptomic data. It is possible that A-to-I editing is somehow restricting viral propagation, thus reducing the number of viral progeny showing evidence of these changes. In Section 2: “Analysis of copy number variations from cell-free DNA of lung cancer patients via Nanopore sequencing” I have developed a customized workflow to exploit Nanopore sequence for the analysis of plasmatic cell-free DNA: Cancer is an extremely dynamic disease: malignant cells are constantly under selective pressure and the evolutionary path of each tumor can take different directions due to such pressure. It is hence important to monitor cancer development at multiple timepoints to closely follow its evolution; unfortunately, the risks and invasiveness of conventional biopsy make it unsuitable for repeated sampling. A valid and non-invasive alternative to tissue sampling is represented by the analysis of cfDNA from liquid biopsy samples (plasma). CNVs are an important class of genetic alterations that can affect tumor aggressivity and resistance to treatment. To date, the only reliable approach to obtain a whole-genome CNV profile from plasmatic cfDNA is Illumina sequencing. However the need for expensive sequencers is often an obstacle for smaller laboratories. Oxford Nanopore Technologies has recently released MinION: a fast and extremely inexpensive third generation sequencer based on the Nanopore technology. However, this technology is not designed for low quality DNA such as cfDNA (very fragmented, low concentration). I have modified Nanopore standard protocols to make them compatible with the characteristics of cfDNA. The technique has been tested on plasma samples obtained from lung cancer patients, with the aim of detecting tumor-specific copy number variations. The approach has been subsequently validated by comparing it with the current standard technique (Illumina). Nanopore and Illumina results strongly correlate (R = 0.96 – 0.99, p << 0.001), with concordant log2ratio values in 97-99% of genome positions. Nanopore features (i.e. reduced costs) represent advantages over current sequencing technologies, and might drive the adoption of molecular karyotyping from liquid biopsies as a tool for cancer monitoring in clinical settings.

Les troubles du spectre autistique (TSA) touchent approximativement 1% de la population générale. Ces troubles se caractérisent par un déficit de la communication sociale, ainsi que des comportements stéréotypés et des intérêts restreints. Plusieurs gènes impliqués dans le déterminisme des TSA ont été identifiés, comme par exemple les gènes NLGN3-4X, NRXN1-3 et SHANK1-3. Au cours des années précédentes, les TSA ont été considérés comme un ensemble complexe de troubles monogéniques. Cependant, les études récentes du génome complet suggèrent la présence de gènes modificateurs (« multiple hits model »). La dyslexie est caractérisée par un trouble dans l’apprentissage de la lecture et de l’écriture qui touche 5--‐15% de la population générale. Les facteurs génétiques impliqués restent pour l’instant inconnus car seuls des gènes ou loci candidats ont été identifiés. Mon projet de thèse avait pour objectif de poursuivre l’identification des facteurs génétiques impliqués dans les TSA et de découvrir un premier facteur génétique pour la dyslexie. Pour cela, deux types de populations ont été étudiés : d’une part des patients atteints de TSA (N>600) provenant de France, de Suède et des Iles Faroe, d’autre part des patients atteints de dyslexie (N>200) provenant de France, en particulier une famille de 11 personnes atteintes sur 3 générations. J’ai utilisé à la fois la technologie des puces à ADN Illumina (600 K et 5M) et le séquençage complet du génome humain pour effectuer des analyses de liaison et d’association. Pour les TSA, grâce aux analyses de CNVs, j’ai pu identifier des gènes candidats pour l’autisme et confirmer l’association de plusieurs gènes synaptiques avec l’autisme. En particulier, l’étude d’une population de 30 patients des îles Faroe a pu confirmer l’implication des gènes NLGN1 et NRXN1 dans l’autisme et identifier un nouveau gène candidat IQSEC3. En parallèle, j’ai exploréPRRT2 localisé en 16p11.2. PRRT2 code pour un membre du complexe SNARE synaptique qui permet la libération des vésicules synaptiques. Je n’ai pas pu mettre en évidence d’association avec les TSA, mais j’ai montré que ce gène important pour certaines maladies neurologiques était sous pression de sélection différente selon les populations. Pour la dyslexie, j’ai effectué une analyse de liaison (méthode des lod-scores) pour une grande famille de 11 individus atteints sur trois générations. Cette étude a permis d’identifier CNTNAP2 comme un gène de vulnérabilité à la dyslexie. Cette découverte est importante car ce même gène est aussi associé aux TSA. Par contre, aucune des 20 variations rares découvertes par le séquençage complet du génome n’est localisée dans les parties codantes du gène. Plusieurs variations localisées dans des régions régulatrices sont candidates. En conclusion, les résultats de ma thèse ont permis d’identifier des gènes candidats pour les TSA, de confirmer le rôle des gènes synaptiques dans ce trouble, de montrer pour la première fois grâce à une analyse de liaison le rôle de CNTNAP2 dans la dyslexie
Autism spectrum disorders (ASD) affect 1% of the general population. These disorders are characterized by deficits in social communication as well as stereotyped behaviors and restricted interests. Several genes involved in the determination of ASD have been identified, such as NLGN3-4, NRXN1-3 and SHANK1-3. In the previous years, ASD have been considered as a complex set of monogenic disorders. Recent studies on the complete genome nevertheless suggest the presence of modifier genes ("multiple hits model"). Dyslexia is characterized by difficulties in learning to read and write. It affects 5-15 % of the general population. Genetic factors involved remain unknown. Only candidate genes or loci have been identified. My thesis had two main objectives: pursuing the identification of genetic factors involved in ASD, and discovering a first genetic factor for dyslexia. I therefore studied two types of populations: on the one hand a group of patients with ASD (N > 600) from France, Sweden and the Faroe Islands, and on the other hand another group of patients with dyslexia (N > 200) from France, and more specifically a family of 11 people followed over 3 generations. I used both Illumina microarrays technology (600K and 5M) and the complete human genome sequencing to conduct linkage and association analyses. Regarding ASD, CNVs (copy number variants) analyses allowed me to confirm the association of several synaptic genes with autism and to identify new candidate genes. In particular, the study of a population of 30 patients from the Faroe Islands confirmed the involvement of NLGN1 and NRXN1 genes in autism and identified a new candidate gene, IQSEC3. At the same time, I explored PRRT2 located in 16p11.2. PRRT2 encodes a member of the synaptic SNARE complex that allows the release of synaptic vesicles. I have not been able to demonstrate any association with ASD, but I showed that this gene, which is important for some neurological diseases, was under different selection pressures according to the population considered. Regarding dyslexia, I realized a linkage analysis (lod-score method) for a large family of 11 individuals, with three generations affected. This study identified the CNTNAP2 gene as a vulnerability factor for dyslexia. This finding is important because this gene is also associated with ASD. Nevertheless, none of the 20 rare variations discovered by whole genome sequencing is localized in the coding parts of the gene. Only several variations localized in regulatory regions are robust candidates. To conclude, my findings enabled the identification of new candidate genes for ASD, the confirmation of the role of synaptic genes in this disorder, and the highlight for the first time of the role of CNTNAP2 in dyslexia through linkage analysis

Ces dix dernières années, l’investigation des maladies génétiques a été bouleversée par l’émergence des techniques de séquençage haut-débit. Celles-ci permettent désormais de ne plus séquencer les gènes un par un, mais d’avoir accès à l’intégralité de la séquence génomique ou transcriptomique d’un individu. La difficulté devient alors d’identifier les variants causaux parmi une multitude d’artefacts techniques et de variants bénins, pour ensuite comprendre la physiopathologie des gènes identifiés.L’application du séquençage haut débit est particulièrement prometteuse dans le champ de la génétique de l’infertilité masculine car il s’agit d’une pathologie dont l’étiologie est souvent génétique, qui est génétiquement très hétérogène et pour laquelle peu de gènes ont été identifiés. Mon travail de thèse est donc centré sur la l’infertilité et comporte deux parties majeures : l’analyse des données issues du séquençage haut débit d’homme infertiles et de modèles animaux et la caractérisation moléculaire d’un phénotype spécifique d’infertilité, laglobozoospermie.Le nombre de variants identifiés dans le cadre d’un séquençage exomique pouvant s’élever à plusieurs dizaines de milliers, l’utilisation d’un outil informatique performant est indispensable. Pour arriver à une liste de variants suffisamment restreinte pour pouvoir être interprétée, plusieurs traitements sont nécessaires. Ainsi, j’ai développé un pipeline d’analyse de données issues de séquençage haut-débit effectuant de manière successive l’intégralité des étapes de l’analyse bio-informatique, c’est-à-dire l’alignement des reads sur un génome de référence, l’appel des génotypes, l’annotation des variants obtenus ainsi que le filtrage de ceux considérés comme non pertinents dans le contexte de l’analyse. L’ensemble de ces étapes étant interdépendantes,les réaliser au sein du même pipeline permet de mieux les calibrer pour ainsi réduire le nombre d’erreurs générées. Ce pipeline a été utilisé dans cinq études au sein du laboratoire, et a permis l’identification de variants impactant des gènes candidats prometteurs pouvant expliquer le phénotype d’infertilité des patients.L’ensemble des variants retenus ont ensuite pu être validés expérimentalement.J’ai également pris part aux investigations génétiques et moléculaires permettant la caractérisation du gène DPY19L2, identifié au laboratoire et dont la délétion homozygote entraine une globozoospermie, caractériséepar la présence dans l’éjaculât de spermatozoïdes à tête ronde dépourvus d’acrosome. Pour cela, j’ai contribué à caractériser les mécanismes responsables de cette délétion récurrente, puis, en utilisant le modèle murin Dpy19l2 knock out (KO) mimant le phénotype humain, j’ai réalisé une étude comparative des transcriptomes testiculaires de souris sauvages et de souris KO Dpy19l2-/-. Cette étude a ainsi permis de mettre en évidence la dérégulation de 76 gènes chez la souris KO. Parmi ceux-ci, 23 sont impliqués dans la liaison d’acides nucléiques et de protéines, pouvant ainsi expliquer les défauts d’ancrage de l’acrosome au noyau chez les spermatozoïdes globozoocéphales.Mon travail a donc permis de mieux comprendre la globozoospermie et de développer un pipeline d’analyse bioinformatique qui a déjà permis l’identification de plus de 15 gènes de la gamétogenèse humaine impliqués dans différents phénotypes d’infertilité
In the last decade, the investigations of genetic diseases have been revolutionized by the rise of high throughput sequencing (HTS). Thanks to these new techniques it is now possible to analyze the totality of the coding sequences of an individual (exome sequencing) or even the sequences of his entire genome or transcriptome.The understanding of a pathology and of the genes associated with it now depends on our ability to identify causal variants within a plethora of technical artifact and benign variants.HTS is expected to be particularly useful in the field infertility as this pathology is expected to be highly genetically heterogeneous and only a few genes have so far been associated with it. My thesis focuses on male infertility and is divided into two main parts: HTS data analysis of infertile men and the molecular characterization of a specific phenotype, globozoospermia.Several thousands of distinct variants can be identified in a single exome, thereby using effective informatics is essential in order to obtain a short and actionable list of variants. It is for this purpose that I developed a HTS data analysis pipeline performing successively all bioinformatics analysis steps: 1) reads mapping along a reference genome, 2) genotype calling, 3) variant annotation and 4) the filtering of the variants considered as non-relevant for the analysis. Performing all these independent steps within a single pipeline is a good way to calibrate them and therefore to reduce the number of erroneous calls. This pipeline has been used in five studies and allowed the identification of variants impacting candidate genes that may explain the patients’ infertility phenotype. All these variants have been experimentally validated using Sanger sequencing.I also took part in the genetic and molecular investigations which permitted to demonstrate that the absence of the DPY192 gene induces male infertility due to globozoospermia, the presence in the ejaculate of only round-headed and acrosomeless spermatozoa. Most patients with globozoospermia have a homozygous deletion of the whole gene. I contributed to the characterization of the mechanisms responsible for this recurrent deletion, then, using Dpy19l2 knockout (KO) mice, I realized the comparative study of testicular transcriptome of wild type and Dpy19l2 -/- KO mice. This study highlighted a dysregulation of 76 genes in KO mice. Among them, 23 are involved in nucleic acid and protein binding, which may explain acrosome anchoring defaults observed in the sperm of globozoospermic patients.My work allowed a better understanding of globozoospermia and the development of a HTS data analysis pipeline. The latter allowed the identification of more than 15 human gametogenesis genes involved in different infertility phenotypes

A infertilidade afeta aproximadamente 15% dos casais, sendo atualmente reconhecido o envolvimento de fatores masculinos em metade dos casos. Alterações nas análises seminais são detectadas na maioria dos homens inférteis e a mais frequente é a baixa concentração de espermatozoides no ejaculado, conhecida como oligozoospermia. Vários estudos mostram uma forte relação entre fatores genéticos e a infertilidade, incluindo alterações cromossômicas e microdeleções do cromossomo Y, porém as causas da oligozoospermia ainda permanecem obscuras. O desenvolvimento de novas tecnologias de investigação vem possibilitando a detecção de alterações a nível genômico, como mutações e variações no número de cópias (CNVs). O presente trabalho teve por objetivo a caracterização genômica de homens com oligozoospermia sem causa definida, visando estabelecer correlação entre alterações no número de cópias e perdas de heterozigosidade (LOHs) e o fenótipo de infertilidade. Foram selecionados 18 pacientes após rigorosa avaliação clínica e investigação do histórico reprodutivo, sendo excluídos pacientes portadores de alterações cromossômicas e portadores de microdeleções do cromossomo Y. Seis homens comprovadamente férteis foram selecionados para o grupo controle. A investigação genômica de ambos os grupos, amostral e controle, foi realizada pela técnica de hibridação genômica comparativa em microarranjos (aCGH) utilizando a plataforma de resolução 180K (Agilent®,US), analisada pelo software Nexus 8.0. Foram detectadas alterações possivelmente patogênicas no cromossomo Y, no cromossomo X e em autossomos. Um ganho na região de AZFc envolvendo apenas os genes DAZ1 e DAZ4 foi detectado em nove pacientes e em quatro controles, sendo classificado como alteração benigna. Porém, alterações na região de AZFc possivelmente relacionadas ao fenótipo de oligozoospermia foram detectadas em três pacientes e incluíram extensas duplicações e deleções envolvendo, entre outros genes, as quatro cópias do gene DAZ. Após comparação de regiões selecionadas com a literatura e com diferentes bancos de dados genéticos, sugerimos que os genes PLEC, SPATC1, COL1A1, MOV10L1, SYCE3 e ODF3B possam estar associados a alterações na produção espermática. Adicionalmente, entre os doze miRNAs presentes em regiões de LOH possivelmente relacionadas ao fenótipo de infertilidade, dez têm como alvo genes com funções relacionadas à espermatogênese e reprodução humana. Estudos adicionais a nível de expressão e sequenciamento gênico são necessários para confirmar a correlação entre o genótipo e o fenótipo de oligozoospermia.
Infertility affects about 15% of the couples, and it is currently recognized, that male factors are involved in about 50% of cases. Changes in seminal parameters are detected in most infertile men and the most common alteration, known as oligozoospermia, is a low concentration of sperm in the ejaculate. Several studies show a strong relationship between genetic factors and infertility, including chromosomal abnormalities and microdeletions of Y chromosome, however, the causes of oligozoospermia remain unclear. The development of new research technologies has allowed the detection of changes at genomic levels, such as mutations and copy number variations (CNVs). This study aimed to perform a genomic characterization of patients with idiopathic oligozoospermia to determine whether there is a correlation between changes of copy number and losses of heterozygosity (LOHs) in relation to the phenotype of infertility. Eighteen patients were selected for the cases after rigorous clinical examination and investigation of their reproductive history. Patients with chromosomal abnormalities or microdeletions of the Y chromosome were excluded. Six proven fertile men comprised the control group. Genomic investigation of both groups was performed by microarray comparative genomic hybridization (aCGH) using 4X180K platform (Agilent, US) analysed by Nexus 8.0 software. Potential pathogenic changes were detected on Y chromosome, as well as on the X and autosome chromosomes. A gain in AZFc region involving only DAZ1 and DAZ4 genes was detected in nine patients and four controls, and was considered as benign. However, changes in AZFc region, that could be related to the oligozoospermia phenotype were detected in three patients. These changes included extensive duplications and deletions involving the four copies of the DAZ gene together with copy number changes affecting other genes. After comparing the selected regions with the literature and with different databases, we suggest that changes such as LOH affecting PLEC, SPATC1, COL1A1, MOV10L1, SYCE3 and ODF3B genes may influence sperm production. Our analysis indicates that, ten out of the twelve miRNAs present in LOH regions could be involved in the infertility phenotype and could have target genes with functions related to spermatogenesis and human reproduction. Additional studies involving gene sequencing and expression analysis are needed to confirm the the correlation between the genotype and oligozoospermia phenotype.

Um eine Entscheidung zu treffen, muss Information interpretiert und in eine Handlung übersetzt werden. Dafür wird die a priori Wahrscheinlichkeit bezüglich der Entscheidungsalternativen in den Prozess der Entscheidungsfindung integriert und löst Mechanismen der Handlungsvorbereitung aus. In der vorliegenden Dissertation habe ich untersucht, welche Vorbereitungsprozesse aufgrund von wahrscheinlichkeitsbasierter Vorinformation stattfinden und welche Gehirnareale mit der Integration dieser Information assoziiert sind. Um diese Fragen zu beantworten, habe ich eine Verhaltensstudie, eine Studie mit Ableitung des Elektroenzephalogramms (EEG) und eine Studie mittels der funktionellen Magnetresonanztomographie (fMRT) mit simultaner Ableitung des EEGs durchgeführt. Die Versuchspersonen bearbeiteten währenddessen eine Zahlenvergleichsaufgabe mit einem Hinweisreiz, der Wahrscheinlichkeitsinformation bezüglich der erforderlichen Antwort enthielt. Die Reaktionszeit wurde durch die wahrscheinlichkeitsbasierte Vorinformation des Hinweisreizes parametrisch moduliert (Studie 1). Daraus lässt sich schlussfolgern, dass Vorbereitungsprozesse in Abhängigkeit der Wahrscheinlichkeitsinformation stattfinden. Die EEG Studie (Studie 2) ergab einen parametrischen Effekt von Wahrscheinlichkeitsinformation auf die Amplitude der Contingent Negative Variation (CNV), einer EEG-Komponente, die Vorbereitungsprozesse auf prämotorischer Ebene reflektiert. Darüber hinaus fand sich mittels einer Dipolquellenanalyse ein Dipol im anterioren Cingulum (ACC), dessen Aktivität ebenfalls durch die Wahrscheinlichkeitsinformation parametrisch moduliert war. Diese Ergebnisse lassen auf prämotorische Vorbereitungsprozesse aufgrund von Wahrscheinlichkeitsinformation schließen. In den fMRT-Ergebnissen zeigte sich eine parametrisch modulierte neuronale Aktivierung im posterioren Teil des medial-frontalen Kortex (pMFC), die auf eine Kontrollfunktion zur Handlungsanpassung dieses Areals zurückgeführt werden kann (Studie 3a). Um dynamische Fluktuationen der Wahrscheinlichkeitsverarbeitung zu untersuchen, wurde die CNV Amplitude der Einzeltrials in das Modell der fMRT-Analyse integriert (Studie 3b). Die CNV Amplitude korrelierte mit der neuronalen Aktivität in einem Netzwerk, bestehend aus frontalen, parietalen und striatalen Arealen, das mit allgemeiner wahrscheinlichkeitsunabhängiger Handlungsvorbereitung im Zusammenhang steht. Dagegen zeigten sich im dorsolateralen Präfrontalkortex (DLPFC), im inferioren frontalen Gyrus (IPG) und im inferioren Parietallappen (IPL) Aktivierungen, die sich auf die dynamische Integration von Wahrscheinlichkeitsinformation zurückführen lassen.
To prepare actions in advance, prior information about the probability of decision alternatives is integrated into the decision-making process. In the present dissertation, I investigated preparatory processes elicited by prior probability (PP) and the neural basis of PP processing. In three studies, I collected behavioral data and, furthermore, recorded electroencephalographic (EEG) data separately as well as simultaneously with functional magnetic resonance imaging (fMRI). While applying these methods, participants had to perform a number comparison task with a precue delivering PP about a subsequent response-demanding stimulus. The probability precue elicited the preparation of the response, as shown by the parametrical modulation of response time (RT) depending on PP (Study 1). The EEG study (Study 2) revealed a parametrical effect of PP on the contingent negative variation (CNV) during the foreperiod, which is an indicator for premotor response preparation. Furthermore, a dipole was located in the anterior cingulate cortex (ACC) with its activity parametrically modulated by PP. These EEG results suggest that PP influences premotor response preparation in a parametrical fashion. An analysis of fMRI data showed that neural activity in the posterior medial frontal cortex (pMFC) increased with increasing PP (Study 3a), which is attributed to a monitoring function of this region with respect to behavioral adjustment and initiation of response preparation depending on the PP. By applying an EEG-informed fMRI analysis (Study 3b), I focused on trial-to-trial fluctuations in PP processing and general response preparation as represented by the single-trial CNV amplitude. I found that the CNV amplitude was correlated with neural activity in a network consisting of frontal, parietal, and striatal regions reflecting general preparatory processes independently of PP. Parts of the network, namely, the dorsolateral prefrontal cortex (DLPFC), the inferior frontal gyrus (IFG), and the inferior parietal lobule (IPL), showed activations, which exclusively represented the contributions of PP to the CNV amplitude fluctuations. These results suggest that PP elicits premotor response preparation and activates the pMFC parametrically signaling the need for behavioral adjustment. In contrast, DLPFC, IFG, and IPL are involved in dynamically fluctuating PP processing mechanisms.

Nella Tesi viene riportata l’analisi genetica di un campione di 128 famiglie con Disturbo dello Spettro Autistico, tramite il sistema di SNP array “PsychArray” (Illumina ), contenente oltre 500.000 sonde sull’intero genoma. Questi dati sono stati utilizzati per individuare Copy Number Variants (CNVs) rari e rilevanti da un punto di vista clinico. Sono stati quindi selezionati tre CNVs per un ulteriore approfondimento: due microdelezioni già descritte come patologiche (rispettivamente nella regione 1p36.32 e 22q13.33 comprendente il gene SHANK3) sono risultate essere “de novo”, mentre una terza microdelezione nel gene CTNNA3 è ereditata dalla madre. Tutti e tre i CNV sono stati validati tramite Real Time-PCR, definendone i confini. Per quanto riguarda la microdelezione in CTNNA3, poiché difetti di questo gene sono stati implicati nell’autismo con un meccanismo recessivo, è stata anche condotta un’analisi di sequenza di tutti gli esoni del gene negli individui della famiglia interessata, al fine di ricercare eventuali mutazioni puntiformi sull’allele non deleto. Questa analisi non ha individuato nessuna variante potenzialmente dannosa, pertanto il difetto in CTNNA3 non risulta essere la causa principale del fenotipo autistico in questa famiglia, anche se potrebbe avere un ruolo come fattore di suscettibilità.

Hintergrund: Zur Begutachtung von Riechschäden werden immer häufiger objektive Befunde benötigt. Die bisher meist übliche Registrierung olfaktorisch evozierter Potenziale (OEP) ist technisch aufwändig und von der Atemtechnik des Probanden abhängig. Zur Diagnostik der Anosmie und Parosmie wird hier die methodisch einfachere Messung der "contingent negative variation" (CNV) eingesetzt. Patienten und Methode: An 25 Probanden mit normalem Riechvermögen und 16 Patienten mit dem subjektiven Befund einer Anosmie nach Unfallverletzung wurden OEP- und CNV-Messungen vorgenommen. Bei der "direkten" CNV sollte der Proband einen Ton aufmerksam erwarten, der einem Duftreiz nach 1,5 s folgte. Für die "selektive" CNV wurde nur einer von zwei zufällig wechselnden Duftreizen mit einem Ton als Zweitreiz markiert, der eine Erwartungsreaktion auslösen sollte. Ergebnisse: Für die beiden Versuchsarten wurde bei 21 bzw. 23 der Probanden mit normalem Riechvermögen eine eindeutige CNV gefunden. Das OEP fehlte in 4,3 % aller Messungen. Bei den Patienten mit Anosmie war in keinem Fall ein OEP bzw. eine CNV vorhanden. Die Amplituden der "selektiven" CNV sind signifikant höher als die der "direkten" CNV. Für die Ergebnisse bei weiblichen und männlichen Probanden ergab sich kein signifikanter Unterschied. Schlussfolgerung: Die Ergebnisse zeigen, dass ein objektiver klinischer Riechtest mit CNV-Ableitung möglich ist. Im Gegensatz zur OEP-Messung, deren Ergebnis von der Reaktion auf die Reizparameter abhängt, ist die CNV ein Maß für die kognitive Bedeutung des Reizes. Die einfache Reiz- und Messtechnik könnte zur weiteren Verbreitung der objektiven Olfaktometrie beitragen.
Background: An objective smelling test is indicated for a reliable assessment of olfactory disorders. Usually olfactory evoked potentials (OEP) are registered. But the technique of this measurement is complicated and the generation of the OEP depends on the respiration of the subject. Alternatively, the contingent negative variation (CNV) can be used in the diagnosis of anosmia and parosmia, requireing only a simple olfactory stimulator. Subjects and Method: OEP and CNV were derived from 25 adults with normal smelling and from 16 patients with anosmia after head injury. First, the "direct" CNV was registered when the subjects expected a tone following a smell stimulus after 1.5 s. Using two different odors in a random order, the tone only followed one of them, so the "selective" CNV was scored. Results: In both tests a distinct CNV was found in 21 and 23 normal smelling subjects, respectively. OEPs were absent in 4.3 % of this control group. No patient with anosmia showed an OEP or a CNV. The amplitudes of the "selective" CNV are significantly higher than those of the "direct" CNV. No gender dependency was found. Conclusion: The results show that an objective olfactometry can be realized by registration of CNV. Contrary to the measurement of OEP which depend on the physical parameters of olfactory stimuli, CNV correlates well with the cognitive identification of odor.

The thesis identify CNV structural variants as possible markers for genomic selection and identify QTL regions for Fatty Acid Content in the Italian Brown Swiss population. Additionally it maps the QTL for mastitis resistance in the Valdostana Red Pied cattle.

The thesis identify CNV structural variants as possible markers for genomic selection and identify QTL regions for Fatty Acid Content in the Italian Brown Swiss population. Additionally it maps the QTL for mastitis resistance in the Valdostana Red Pied cattle.

Au cours des 20 dernières années, l'évolution des nouvelles technologies a révélé la grande variabilitéde notre génome depuis la simple substitution jusqu'aux réarrangements chromosomiques. Lestechnologies de séquençage à haut débit ont particulièrement amélioré l’identification etl’interprétation des variations de petite taille tout en offrant l’opportunité d’explorer les variations destructure avec une résolution supérieure à celle disponible grâce aux analyses pangénomiques surpuces. Néanmoins, l’identification des variations de structure, et plus particulièrement des variationsdu nombre de copies (CNV) à partir de données de séquençage par capture, a été sous exploitée etpeu évaluée. Notre objectif principal était de mettre en place un pipeline bioinformatique basé sur laprofondeur de lecture pour l’identification des CNV, puis de l’appliquer à une études cas-témoinsd’exome dans le cadre de la recherche sur la maladie d’Alzheimer.La maladie d’Alzheimer (MA) est la maladie neurodégénérative la plus fréquente. Les facteursgénétiques individuels jouent un rôle important dans son déterminisme et de multiples facteurs derisque ont été identifiés, essentiellement des substitutions et petites insertions/délétions. Pourtant,des variations de structure ont déjà été identifiées dans des formes monogéniques de MA, comme lesduplications complètes du gène APP. Les CNV restent très peu étudiés dans la MA et nous avonssouhaité appliquer une approche cas-témoins à partir de données massives d’exomes pour détecterdes CNV contribuant au risque de MA.Dans un premier temps, nous avons établi une stratégie d'analyse basée sur le logiciel CANOES afin dedétecter les CNV à partir de données de NGS issues d’une capture (panel, exomes). Cette approche aété validée à travers deux grands jeux de données de panels et d’exomes comparés à des techniquesindépendantes. Dans le premier jeu de données (panels), la sensibilité et la spécificité étaient de 100%et nous obtenons une sensibilité de 87,25 % et une valeur prédictive positive de 88,5% sur la détectionde CNV sur les données de séquençage d'exomes.Par la suite, nous avons appliqué cette approche aux données d’exomes issues des consortium ADES(Alzheimer Disease Exome Sequencing) et ADSP (Alzheimer Disease Sequencing Project), regroupant,après un contrôle qualité extensif développé dans le cadre de ces travaux, 22 094 individus répartisentre 4077 formes précoces de MA, 8458 formes tardives et 9559 témoins. Nous avons mis au pointdes analyses au niveau des transcrits et appliqué une méthode statistique basée sur les dosagesappliquée aux formes précoces et aux témoins. Nous avons pu identifier plusieurs potentiels nouveauxfacteurs de risque dont la région du chr22q11.21, déjà impliquée dans les troubles duneurodéveloppement (p=3,8x10-4). De plus, nous avons identifié des délétions très rares dans lesgènes ABCA1 et ABCA7 dont les variations perte de fonction sont connues comme facteurs de risquede MA depuis peu, et nous avons réalisé une analyse conjointe des délétions et des variations pertede fonction de petite taille.En conclusion, nous avons montré que la détection de CNV issus de données d’exome est fiable et nousen avons mesuré les performances et les limites avant de les appliquer à un grand jeu de données afind’identifier de nouveaux mécanismes contribuant au développement de la maladie d’Alzheimer
Over the past 20 years, the evolution of new technologies has revealed the great variability of ourgenome, from simple substitutions to chromosomal rearrangements. High-throughput sequencing hasparticularly improved the identification and interpretation of small variations, while offering theopportunity to explore structural variations with a higher resolution than that available with genome-wide microarray analyses. Nevertheless, the identification of structural variations and more specificallycopy number variations (CNVs) from capture sequencing data, has been under exploited and underevaluated. Our main objective was to develop a read depth based bioinformatics pipeline for CNVidentification, and then apply it to a case-control exome study in Alzheimer’s disease research.Alzheimer’s disease (AD) is the most common neurodegenerative disorder. Individual genetic factorsplay an important role in its determinism, and multiple risk factors have been identified, mainlysubstitutions and small insertions/deletions. However, structural variations have already beenidentified in monogenic forms of AD, such as complete duplication of APP gene. CNVs remain largelyunstudied in AD, we set out to apply a case-control approach using massive exome data to detect CNVscontributing to AD risk.As a first step, we established an analysis strategy based on CANOES software to detect CNVs fromNGS data derived from a capture (gene panels, exomes). This approach was validated using 2 largegene panels and exome datasets, compared with independent targeted techniques. In the first dataset(gene panels), sensitivity and specificity were 100%, and we obtained a sensitivity of 87.25% and apredictive positive value of 88.5% for CNV detection in whole exome sequencing data.We then applied this approach to whole exome data from the ADES (Alzheimer Disease ExomeSequencing) and ADSP (Alzheimer Disease Sequencing Project) consortia, grouping, after extensivequality control developed as part of this work, 22,094 samples divided between 4077 early onset cases,8458 late onset and 9559 controls. We developed transcript-level analyses and applied a statisticalmethod based on dosage applied on early onset cases and controls. We were able to identify severalpotential new risk factors, including the 22q11.21 regions, already implicated in neurodevelopmentaldisorders (p=3,8x10-4). In addition, we identified rare deletions in ABCA1 and ABCA7 genes, whoseloss-of-function variations have recently been identified as risk factors for AD, and carried out a jointanalysis of deletions and small loss-of-function variations.In conclusion, we have shown that CNV detection from exome data is reliable, and we have measuredits performance and limitations before applying it to a large dataset to identify new mechanismscontributing to the development of Alzheimer's disease

O uso de ferramentas de análise e visualização de dados é essencial para a pesquisa de CNVs, porém nem sempre está ao alcance de todo o meio científico devido a restrições de acesso ou por requerer conhecimento avançado de informática. Portanto o desenvolvimento de interfaces amigáveis e acessíveis é essencial para a pesquisa. Esta dissertação visou explorar o ambiente dos navegadores Web para desenvolver soluções para os problemas de acessibilidade, portabilidade e visualização, comumente encontrados nas ferramentas de bioinformática. Foi desenvolvido um aplicativo para análise exploratória, denominado CNViewer, o qual oferece recursos para comparação de perfis moleculares, além de representar graficamente diversas amostras simultaneamente. Através de uma interface dinâmica, o usuário pode delimitar quaisquer regiões genômicas para a análise, e pode também exibir dados complementares às CNVs. Foi também disponibilizado acesso direto a anotações genômicas, tornando o CNViewer um ambiente para estudo de CNVs e dados correlacionados. Utilizando somente os recursos oferecidos pelos navegadores Web (JavaScript e HTML), o CNViewer é capaz de processar dados e executar tarefas rapidamente, com independência de servidor, pois mantêm os dados em memória durante seu uso, aperfeiçoando a interatividade com o usuário. Foi também criado um módulo de exportação, que permite ao usuário salvar e recuperar suas análises, servindo também para compartilhamento de dados. O CNViewer é um aplicativo que supera os limites dos programas baseados em Web clássicos, comportando-se como um aplicativo desktop, mas com a vantagem de ser acessado diretamente, sem requerer instalação ou atualização. O crescente uso dos navegadores Web como ambiente de trabalho, e mais recentemente até mesmo como sistema operativo, sugere que aplicativos nativos a esse ambiente poderão tornar-se a norma em informática biomédica.

Fibromyalgia (FM) is a highly disabling syndrome defined by a low pain threshold and a permanent state of pain. Widespread pain is accompanied by a constellation of symptoms such as fatigue, sleep disturbances and cognitive impairment, among others. The mechanisms explaining this chronic pain remain unclear. Nowadays, the most established/ plausible hypothesis underlying FM ethiopathogenesis is the existence of a dysfunction in pain processing, as supported by alterations in neuroimaging and neurotransmitters levels. The etiology of FM involves the interaction of environmental and genetic susceptibility factors. The genetic contribution to FM has been proven by the presence of a higher concordance of monozygotic than dizygotic twins as well as family aggregation. However, the individual genetic and environmental factors involved have not been identified. The aim of this thesis was to elucidate genetic susceptibility factors for fibromyalgia. We assessed this objective through three main approaches: the identification of FM clinically homogeneous subgroups with a two step cluster analyses, a genome-wide association study in order to evaluate the possible contribution of single nucleotide polymorphisms with Illumina 1 million duo array, and array comparative genomic hybridization experiments to identify regions varying in copy number that could be involved in FM susceptibility,using Agilent 2X400K platform. 48 variables were evaluated in 1,446 Spanish FM cases fulfilling 1990 ACR FM criteria. A partitioning analysis was performed to find groups of variables similar to each other. Variables clustered into three independent dimensions: “symptomatology”, “comorbidities” and “clinical scales”. Only the two first dimensions were considered for the construction of FM subgroups, classifying FM samples into three subgroups: low symptomatology and comorbidities (Cluster 1), high symptomatology and comorbidities (Cluster 2), and high symptomatology but low comorbidities (Cluster 3). These subgroups showed differences in measures of disease severity and were further implemented in genetic analysis. Genome-wide association study was performed in 300 FM cases and 203 controls. No SNP reached GWAS association threshold, but 21 of the most associated SNPs were chosen for replication in over 900 cases and 900 pain free-controls. Four of the strongest associated SNPs selected for replication showed a nominal association in the joint analysis. In particular, rs11127292 (MYT1L) was found to be associated to FM with low comorbidities. Array comparative genomic hybridization detected 5 differentially hybridized regions. They were followed up and one of these regions was validated though a multiplex PCR experiment. An intronic deletion in NRXN3 showed to be associated to female cases of FM and in particular those with low levels of comorbidities. Replication analysis showed a stronger association when considering only female cases and controls and low comorbidities. This enhance the importance of gender in FM etiopathogenesis and could be pointing to the existence of a different genetic background for FM in males and females highlights the importance of identifying FM homogeneous subgroups for the detection of FM genetic susceptibility factors. If the proposed FM candidate genes are further validated in replication studies, this would constitute a change in the FM ethiologycal concept, as several of these candidates are known neuropsychiatric disease associated genes (autism, addiction, mental disability). This would highlight a novel neurocognitive involvement in this disorder, currently considered musculoskeletal and affective.
La fibromialgia (FM) es una enfermedad de etiología desconocida que se caracteriza por dolor crónico generalizado, junto a una amplia constelación de síntomas acompañantes. La base etiopatogénica que explica este estado permanente de dolor es aún desconocida. Hasta la fecha la teoría más plausible es la existencia de una disfunción en la transmisión del dolor. Los estudios familiares han mostrado una considerable agregación familiar en FM,sugiriendo la importancia de los factores genéticos en el desarrollo de estos cuadros. Con la presente tesis se ha pretendido estudiar e identificar variantes del genoma (polimorfismos de base única (SNPs)- y variantes en el número de copia –CNVs) asociadas a FM, con el objetivo de profundizar en la etiología de la enfermedad. Para ello, se han llevado a cabo tres grandes aproximaciones: la identificación de subgrupos clínicos homogéneos de FM mediante un análisis de clusters, un estudio de genoma completo (GWAS) para el análisis directo de SNPs y experimentos de hibridación genómica comparada mediante arrays (aCGH) con el fin de identificar regiones variables en el número de copia asociadas a FM. El análisis de clusters ha permitido la identificación de tres subgrupos de FM en función de los niveles de sintomatología y de comorbilidad personal y familiar. Los resultados del GWAS indican una posible contribución del sistema nervioso central en el desarrollo de FM, ya que las enfermedades neurológicas aparecen como sobrerrepresentadas en el estudio de pathways realizado en los SNPs que presentaban mayor asociación, y un SNP en el gen MYT1L ha presentado asociación estadísticamente significativa con FM con niveles bajos de comorbilidad, poniendo de manifiesto la importancia de identificar subgrupos clínicamente homogéneos para la detección de factores de susceptibilidad genética para FM. Un CNV en el gen neurexina3 ha mostrado, asimismo, asociación en mujeres con FM, y en particular, en aquellas con bajos niveles de comorbilidad, siendo un nuevo argumento a favor de la implicación del SNC. La confirmación de las variantes detectadas en nuevas cohortes de fibromialgia supondría un giro conceptual de la enfermedad hacia una visión más neurocognitiva que osteomuscular.

L’étiologie des troubles du spectre autistique (TSA) est multifactorielle mettant en jeu une forte composante génétique ainsi que des facteurs environnementaux. A ce jour, il n’existe aucun biomarqueur facilitant un diagnostic précoce des TSA, ce qui permettrait une meilleure prise en charge des patients. De plus, les mécanismes physiopathologiques sont encore mal connus. La majorité des études sur les TSA ont été réalisées sur des populations occidentales. Ainsi, nous avons voulu étudier un groupe de patients libanais atteints de TSA. Nos objectifs s’articulent autour de trois volets : 1) L’identification de facteurs de risques environnementaux dans le but de mieux comprendre la pathologie. Les résultats de cette étude ont montré que plusieurs facteurs tels que la consanguinité, le stress au cours de la grossesse et la prématurité semblent être des facteurs de risques. 2) L’identification des variations structurales dans le génome des enfants atteints de TSA afin d'évaluer la contribution de CNV par l'approche de CGH array à haute résolution. Cette étude nous a permis d’identifier de nouveaux gènes dans les TSA ainsi de confirmer la présence des régions et des gènes déjà cités dans des études précédentes sur les TSA. 3) L’identification des métabolites urinaires et des voies métaboliques associés afin de proposer des biomarqueurs aidant à un diagnostic précoce et à une meilleure compréhension de la physiopathologie de la maladie. De nouveaux métabolites ont été identifiés, comme nous avons confirmé des variations anomalies déjà citées dans la littérature concernant certains métabolites qui pourraient représenter des marqueurs robustes de la maladie
The etiology of autism spectrum disorders (ASD) is multifactorial involving a strong genetic component as well as environmental factors. To date, there is no biomarker facilitating early diagnosis of ASD to improve patient management and outcomes. In addition, the physiopathological mechanisms are still poorly understood. Most studies on ASD have been conducted in Western populations. Thus, we wanted to study a group of Lebanese patients with ASD. Our objectives focus around three aspects: 1) The identification of environmental risk factors in order to better understand the pathology. The results of this study showed that several factors such as consanguinity, stress during pregnancy and prematurity appear to be risk factors. 2) The identification of structural variations in the genome of children with ASD in order to evaluate the contribution of CNV by high resolution CGH array approach. This study allowed us to identify new genes in ASD as well as to confirm the presence of regions and genes already cited in previous ASD studies. 3) The identification of metabolites and associated metabolic pathways to provide biomarkers for early diagnosis and to better understanding of the pathophysiology of the disease. New metabolites have been identified and we have confirmed variations already mentioned in the literature concerning certain metabolites that could represent robust markers of the disease