Orientador: David Corrêa Martins Júnior / Dissertação (mestrado) - Universidade Federal do ABC, Programa de Pós-Graduação em Neurociência e Cognição, 2016. / A biologia sistêmica é um campo de pesquisa interdisciplinar que estuda as complexas interações que ocorrem entre os componentes biológicos de um organismo vivo com o objetivo de entender o seu comportamento, o qual emerge a partir dessas interações. Essas interações compõem uma rede altamente complexa, cujos interagentes podem ser de diversas naturezas. Nesse contexto, as doenças complexas são caracterizadas justamente por serem poligênicas e multifatoriais, ou seja, a gênese e o desenvolvimento dessas doenças são uma consequência da interação conjunta de diversos fatores, incluindo não apenas genes, proteínas e outras moléculas, como também fatores epigenéticos e ambientais. No entanto, diferentes métodos de priorização gênica apresentam resultados (listas de genes) com baixa convergência. Assim, a comparação desses métodos é uma questão crucial. Os objetivos principais da presente dissertação foram a realização de uma extensa revisão da literatura em relação às técnicas de priorização de genes associados a doenças complexas e a comparação de algumas dessas técnicas. Foram selecionadas duas ferramentas: o WGCNA (Weighted Gene Correlation Network Analysis) e o NERI (Network-Medicine Relative Importance), ambos métodos que baseiam-se em teoria de redes complexas e co-expressão para priorização gênica, sendo que o NERI tem o diferencial de modelar as hipóteses da Network Medicine para priorização com base na integração de dados de expressão, de redes de interação proteína-proteína (PPI) e de estudos de associação. Para comparação dos resultados, foram utilizados três bancos de dados de expressão gênica relacionados a esquizofrenia. Como previsto, devido ao diferencial de integração de dados proposto pelo NERI, tal técnica resultou em listas de genes com replicação superior à obtida pelo WGCNA para os três bancos de dados em questão. Além disso a interseção entre as listas de genes priorizados de cada metodologia foi baixa, com poucos genes sendo compartilhados pelos resultados dos dois métodos. Ambas metodologias selecionaram genes com relevância biológica relacionada a esquizofrenia, incluindo grupos de genes relacionados a atividade do sistema imune (infecções, estresse), atividade do Sistema Nervoso Central (atividade sináptica, crescimento axonal) e também de embriogênese. Baseando-se nesses resultados, conclui-se que a análise de redes e a integração de dados biológicos são fundamentais para uma ferramenta apresentar resultados promissores, sobretudo no âmbito da descoberta de novos genes e suas redes de interação biológica que seriam possivelmente desconhecidas se fosse realizada apenas a análise individual de cada tipo de dado biológico disponível. / Systems Biology is an interdisciplinary research field which studies the complex interactions that occur between biological compounds of a living organism in order to understand their behavior, which emerges from these interactions. Such interactions compose a highly complex network, whose elements can be of several types. In this context, complex diseases are characterized precisely by being of polygenic and multifactorial nature, i.e., the genesis and development of these diseases are a result of the joint interaction of several factors, including not only genes, proteins and other molecules, but also epigenetic and environmental factors. However, many methods for gene prioritization present results (list of genes) with small convergence. Thus, the comparison involving those methods is a crucial issue. The main objectives of this master thesis was to perform an extensive literature review related to gene prioritization techniques associated to complex diseases and the comparison of part of these techniques. Two techniques were selected: WGCNA (Weighted Gene Correlation Network Analysis) and NERI (Network-Medicine Relative Importance), both methods based on complex networks theory and co-expression for gene prioritization, but NERI having the differential of modeling the Network Medicine hypotheses for prioritization based on integration of expression, protein-protein interaction (PPI) network and association studies. For comparison of the results, three gene expression databases related to schizophrenia were adopted. As predicted, due to the data integration proposed by NERI, such technique resulted in genes lists with superior replication for the three databases mentioned. Additionally, the intersection between the results of the genes lists prioritized by the two methodologies was small, with few genes being found in both lists. Both methods selected biologically relevant to schizophrenia, including groups of genes related to imune system activity (infections, stress), Central Nervous System activity (synaptic activity, axonal growth) and embryogenesis. From these results, it follows that network analysis and biological data integration are fundamental for a gene prioritization method to present promising results, mainly for discovery of new genes and their biological interaction networks that would possibly be unknown if only an individual analysis of each biological data available were performed.

Hintergrund: Die Berlin Fat Mouse Inzuchtlinie (BFMI) ist ein Modell für Adipositas und das metabolische Syndrom. Diese Studie zielte darauf ab, genetische Varianten zu identifizieren, die mit dem gestörten Glukosestoffwechsel assoziiert sind, indem die fettleibigen Linien BFMI861-S1 und BFMI861-S2 verwendet wurden, die genetisch eng verwandt sind, sich aber in mehreren Merkmalen unterscheiden. BFMI861-S1 ist insulinresistent und speichert ektopisches Fett in der Leber, während BFMI861-S2 insulinsensitiv ist. Methoden: Die QTL-Analyse wurde in zwei fortgeschrittenen Intercross-Linien (AIL) in der Generation durchgeführt. Eine AIL wurde aus der Kreuzung BFMI861-S1 x BFMI861-S2 und die zweite AIL aus der Kreuzung BFMI861-S1 x BFMI861-B6N erhalten. Für beide AILs wurden Phänotypen über 25 bzw. 20 Wochen gesammelt. Zur Priorisierung von positionellen Kandidatengenen wurden Gesamtgenomsequenzierung und Genexpressionsdaten der Elternlinien verwendet. Ergebnisse: Für den AIL BFMI861-S1 x BFMI861-S2 wurden überlappende QTL für das Gonadenfettgewebegewicht und die Blutglukosekonzentration auf Chromosom (Chr) 3 (95,8–100,1 Mb) und für das Gonadenfettgewebegewicht, Lebergewicht und Blut nachgewiesen Glukosekonzentration auf Chr 17 (9,5–26,1 Mb). Für die AIL BFMI861-S1 x BFMI861-B6N zeigte ein hochsignifikanter QTL auf Chromosom (Chr) 1 (157–168 Mb) einen Zusammenhang mit dem Lebergewicht. Ein QTL für das Körpergewicht nach 20 Wochen wurde auf Chr 3 (34,1 – 40 Mb) gefunden, der sich mit einem QTL für das scAT-Gewicht überlappte. In einem multiplen QTL-Mapping-Ansatz wurde ein zusätzliches QTL, das das Körpergewicht bei 16 Wochen beeinflusste, auf Chr 6 (9,5–26,1 Mb) identifiziert. Schlussfolgerungen: Die QTL-Kartierung zusammen mit einem detaillierten Priorisierungsansatz ermöglichte es uns, Kandidatengene zu identifizieren, die mit Merkmalen des metabolischen Syndroms in beiden AILs assoziiert sind. / Background: The Berlin Fat Mouse Inbred line (BFMI) is a model for obesity and the metabolic syndrome. This study aimed to identify genetic variants associated with the impaired glucose metabolism using the obese lines BFMI861-S1 and BFMI861-S2, which are genetically closely related, but differ in several traits. BFMI861-S1 is insulin resistant and stores ectopic fat in the liver, whereas BFMI861-S2 is insulin sensitive. Methods: QTL-analysis was performed in two advanced intercross lines (AIL) in generation. One AIL obtained from the cross BFMI861-S1 x BFMI861-S2 and the second AIL from the cross BFMI861-S1 x BFMI861-B6N. For both AILs phenotypes were collected over 25 and 20 weeks, respectively. For prioritization of positional candidate genes whole genome sequencing and gene expression data of the parental lines were used. Results: For the AIL BFMI861-S1 x BFMI861-S2 overlapping QTL for gonadal adipose tissue weight and blood glucose concentration were detected on chromosome (Chr) 3 (95.8-100.1 Mb), and for gonadal adipose tissue weight, liver weight, and blood glucose concentration on Chr 17 (9.5-26.1 Mb). For the AIL BFMI861-S1 x BFMI861-B6N one highly significant QTL on chromosome (Chr) 1 (157–168 Mb) showed an association with liver weight. A QTL for body weight at 20 weeks was found on Chr 3 (34.1 – 40 Mb) overlapping with a QTL for scAT weight. In a multiple QTL mapping approach, an additional QTL affecting body weight at 16 weeks was identified on Chr 6 (9.5-26.1 Mb). Conclusions: QTL mapping together with a detailed prioritization approach allowed us to identify candidate genes associated with traits of the metabolic syndrome in both AILs.