The nucleosome remodeling factor (NURF) is an evolutionary conserved ATP-dependent chromatin remodeling factor. It was first isolated from Drosophila as a complex with enzymatic activity that once recruited to nucleosome, it slides the nucleosome to provide accessibility for transcription factors. Since then, numerous works from animal models and cell lines show the role of NURF as a regulator of gene expression. NURF interacts with H3K4me3 and sequence specific transcription factors that recruit the complex to promoter regions. Whether this is the only mechanism by which NURF regulates gene expression is not known. However, other ATP-dependent chromatin remodeling complexes are known to regulate gene expression independent from transcription initiation. In order to explore the role of NURF in regulating gene expression, we utilized two genome wide approaches to map NURF binding and NURF dependent changes in chromatin structure using ChIP-Seq and FAIRE-Seq, respectively. From these analyses, we discovered that NURF broadly localizes in the genome with preferences to gene bodies and 3’ends of genes. Also, we found that NURF maintains open chromatin regions at upstream, intron and downstream of genes. These novel findings shed light on new roles for NURF complex within genes, in addition to its classical role at promoter regions. Furthermore, we discovered the function of a previously uncharacterized domain in the NURF specific subunit BPTF. We show that the N-terminal the plant homeodomain (PHD) of BPTF directly interacts with THOC4, a protein associated with RNA-pol 2. Also, we show using ChIP analyses that this interaction recruits BPTF to gene bodies. Next, we investigated functional consequences for NURF recruitment to gene bodies using Cyclin D1 (Ccnd1) gene as a model. These analyses revealed that NURF is required for normal mRNA processing and loss of NURF induces intron retention, which results in unstable transcripts. Finally, we show that the defect in mRNA processing is not specific to the Ccnd1 gene, as we observe similar defects in four other BPTF dependent genes. Together, our work uncovered new role of mammalian NURF complex in regulating gene expression through mRNA processing.

Le cancer du côlon (CC) est le 3ème cancer le plus fréquent en France avec plus de 40 000 nouveaux cas diagnostiqués par an. Vingt-cinq à 40% des CC sont diagnostiqués à un stade II. La majorité des CC de stade II ne nécessite pas de traitement adjuvant par chimiothérapie, cependant, environ 20% de ces patients vont évoluer avec l'apparition de métastases viscérales. Or, seuls les facteurs histopathologiques ou la détermination du statut avec instabilité des microsatellites (type MSI) permettent d’orienter ou non vers une chimiothérapie complémentaire à la chirurgie. La majorité des CC de stade II ont un profil avec stabilité des microsatellites (type MSS) et les facteurs pronostiques actuels sont mal définis. La découverte de marqueurs pronostiques pour ces patients est donc un enjeu majeur de la prise en charge du CC. De cette façon, notre équipe a mis en évidence l’existence d’une signature moléculaire sur 166 tumeurs, portant sur l’expression différentielle de 7 gènes, entre les tumeurs de CC de stade II et les tumeurs de CC de stade III. Ensuite, le statut mutationnel des gènes KRAS et BRAF, effectué sur 803 CC métastatiques, a mis en évidence que la mutation BRAF V600E est principalement associée au CC de type MSS suggérant un impact pronostique négatif. Enfin, l’étude du statut du gène APC (mutations, LOH, hyperméthylation du promoteur) sur 183 CC de stade II montre que ce gène n’est vraisemblablement pas impliqué dans le développement de métastases du CC. En perspective de ce travail, l’arrivée de techniques de séquençage de nouvelle génération permet d’envisager un traitement adapté à la tumeur, orientant vers une médecine personnalisée. / Colon cancer (CC) is the third most common cancer in France with 40000 new cases diagnosed every year. For thirty years, death-rate has decreased through better therapeutic and screening management, and 40% of CC are diagnosed of stage II. Most of stage II CC do not require chemiotherapy adjuva,t to surgical resection. About 20% of patients with stage II CC relapse within 5 years following the diagnosis. Except microsatelitte instability (MSI) incombination with few hispathologic parameters, the molecular prognosis factors are not well defined and remain a major biological challenge in stage II CC. Our study analyzed the molecular signature of 166 tumors and determined the different expression of seven genes between stage II and stage III CC. KRAS and BRAF mutations were determined on 803 metastatic CC showing an association between V600E BRAF mutation and tumors with microsatellite stability (MSS). This result suggests a negative prognostic impact of BRAF mutation in MSS CC. Finally, the APC gene statut (mutations, LOH, promoter hypermethylation) of 183 stage II CC shows that this gene is probably not involved in the metastatic process. Further developments will consist in applying the next generation sequencing to allow the simultaneous analysis of hundreds target genes. In this way, the treatment will be adapted to each tumor, moving towards personalized medicine.

Cette thèse aborde deux problèmes relatifs à l’analyse et au traitement des données biologiques à haut débit: le premier touche l’analyse bioinformatique des génomes à grande échelle, le deuxième est consacré au développement d’algorithmes pour le problème de la recherche d’un classement consensus de plusieurs classements.L’épissage des ARN est un processus cellulaire qui modifie un ARN pré-messager en en supprimant les introns et en raboutant les exons. L’hétérodimère U2AF a été très étudié pour son rôle dans processus d’épissage lorsqu’il se fixe sur des sites d’épissage fonctionnels. Cependant beaucoup de problèmes critiques restent en suspens, notamment l’impact fonctionnel des mutations de ces sites associées à des cancers. Par une analyse des interactions U2AF-ARN à l’échelle génomique, nous avons déterminé qu’U2AF a la capacité de reconnaître environ 88% des sites d’épissage fonctionnels dans le génome humain. Cependant on trouve de très nombreux autres sites de fixation d’U2AF dans le génome. Nos analyses suggèrent que certains de ces sites sont impliqués dans un processus de régulation de l’épissage alternatif. En utilisant une approche d’apprentissage automatique, nous avons développé une méthode de prédiction des sites de fixation d’UA2F, dont les résultats sont en accord avec notre modèle de régulation. Ces résultats permettent de mieux comprendre la fonction d’U2AF et les mécanismes de régulation dans lesquels elle intervient.Le classement des données biologiques est une nécessité cruciale. Nous nous sommes intéressés au problème du calcul d’un classement consensus de plusieurs classements de données, dans lesquels des égalités (ex-aequo) peuvent être présentes. Plus précisément, il s’agit de trouver un classement dont la somme des distances aux classements donnés en entrée est minimale. La mesure de distance utilisée le plus fréquemment pour ce problème est la distance de Kendall-tau généralisée. Or, il a été montré que, pour cette distance, le problème du consensus est NP-difficile dès lors qu’il y a plus de quatre classements en entrée. Nous proposons pour le résoudre une heuristique qui est une nouvelle variante d’algorithme à pivot. Cette heuristique, appelée Consistent-pivot, s’avère à la fois plus précise et plus rapide que les algorithmes à pivot qui avaient été proposés auparavant. / It is thought to be more and more important to solve biological questions using Bioinformatics approaches in the post-genomic era. This thesis focuses on two problems related to high troughput data: bioinformatics analysis at a large scale, and development of algorithms of consensus ranking. In molecular biology and genetics, RNA splicing is a modification of the nascent pre-messenger RNA (pre-mRNA) transcript in which introns are removed and exons are joined. The U2AF heterodimer has been well studied for its role in defining functional 3’ splice sites in pre-mRNA splicing, but multiple critical problems are still outstanding, including the functional impact of their cancer-associated mutations. Through genome-wide analysis of U2AF-RNA interactions, we report that U2AF has the capacity to define ~88% of functional 3’ splice sites in the human genome. Numerous U2AF binding events also occur in other genomic locations, and metagene and minigene analysis suggests that upstream intronic binding events interfere with the immediate downstream 3’ splice site associated with either the alternative exon to cause exon skipping or competing constitutive exon to induce inclusion of the alternative exon. We further build up a U2AF65 scoring scheme for predicting its target sites based on the high throughput sequencing data using a Maximum Entropy machine learning method, and the scores on the up and down regulated cases are consistent with our regulation model. These findings reveal the genomic function and regulatory mechanism of U2AF, which facilitates us understanding those associated diseases.Ranking biological data is a crucial need. Instead of developing new ranking methods, Cohen-Boulakia and her colleagues proposed to generate a consensus ranking to highlight the common points of a set of rankings while minimizing their disagreements to combat the noise and error for biological data. However, it is a NP-hard questioneven for only four rankings based on the Kendall-tau distance. In this thesis, we propose a new variant of pivot algorithms named as Consistent-Pivot. It uses a new strategy of pivot selection and other elements assignment, which performs better both on computation time and accuracy than previous pivot algorithms.

De nombreux modèles ont été proposés pour expliquer les mécanismes de développement et de progression tumorale, néanmoins certains aspects comme la nature et la hiérarchie des altérations primaires et secondaires sont encore discutés. Pour répondre à ces questions, nous nous sommes intéressés à la progression tumorale des tumeurs hypophysaires à prolactine humaines. Ces tumeurs d'origine monoclonale sont souvent bénignes mais certaines présentent un phénotype agressif voire malin. Afin d'identifier les mécanismes impliqués dans la progression vers le phénotype agressif nous avons utilisé des techniques de génomique intégrative (puces à ADN, séquençage haut débit) couplant l'étude du transcriptome, du génome (variation du nombre de copie chromosomique, polymorphismes) et du miRNome à partir des mêmes échantillons tumoraux. Par cette stratégie nous avons identifié et hierarchisé des altérations spécifiques des tumeurs agressives et / ou malignes dans un modèle expliquant la progression tumorale des tumeurs hypophysaires à prolactine humaines. Nous avons montré que la sous-expression des microARNs miR-183, miR-744 et miR-98 stimule la prolifération via la surexpression de leurs cibles KIAA0101, TGFB1 et E2F2 spécifiquement dans les tumeurs agressives. Ceci entraine l'acquisition d'altérations chromosomiques (perte du chromosome 11 et le gain du chromosome 1q) permettant l'activation de la dissémination métastatique. Enfin, ce travail montre que l'approche de génomique intégrative multidimensionnelle permet d'apporter de nouveaux éléments pour la caractérisation des phénotypes tumoraux, le diagnostic des tumeurs agressives et la prédiction du comportement tumoral / Numerous models have been proposed to explain the mechanisms of tumor development and progression. Nevertheless some aspects such as the nature and hierarchy of primary and secondary alterations are still debated. To answer these questions, we decided to focus on the tumoral progression of human prolactin pituitary tumors. These monoclonal tumors are usually benign but some present an aggressive or malignant phenotype. To identify the molecular events involved in tumoral progression of human PRL towards aggressive phenotype we used an integrative genomics approach (microarrays, high-throughput sequencing) coupling analysis of transcriptome, genome (variation in the number of chromosomal copy polymorphisms) and miRNome from the same human tumor. Using this strategy we identified and prioritized specific alterations of aggressive and / or malignant tumors in a model explaining the tumor progression of human prolactin pituitary tumors. We have shown that under-expression of micro-RNA miR-183, miR-744 and miR-98 stimulates proliferation through overexpression of their targets KIAA0101, TGFB1 and E2F2 specifically in aggressive tumors. This leads to the acquisition of chromosomal damage (loss of chromosome 11 and gain of chromosome 1q) which allowed the activation of the metastatic processes. Finally, this work shows that the integrative genomic multi-dimensional approach can provide new clues for the characterization of tumor phenotypes, diagnosis of aggressive tumors and prediction of tumor behavior

Orientador: Wilter Ricardo Russiano Vicente / Coorientador: Alexeia Barufatti Grisolia / Coorientador: José Fernando Garcia / Banca: Roberto Carvalheiro / Banca: Joaquim Mansano Garcia / Banca: Fernando Sebastián Baldi Rey / Banca: Lenira El Faro / Resumo: Análises de associação genômica ampla (GWAS) para idade ao primeiro parto (IPP) e duração do período gestacional (PG) em bovinos da raça Nelore foram conduzidas utilizando painel de ~ 777.000 polimorfismos de nucleotídeo de sítio único (SNP). As análises de associação foram conduzidas em três grupos amostrais distintos com seus respectivos dados fenotípicos: 1) IPP em meses para 96 fêmeas, 2) valores genéticos preditos (EBVs) para IPP e PG para 831 touros, e 3) EBVs para IPP e PG para 1.278 fêmeas. Para cada conjunto de dados e para cada marcador, efeitos de substituição alélica foram estimados por meio do método dos quadrados mínimos ponderados e corrigidos para estratificação populacional. Os marcadores apresentando maiores efeitos estimados foram considerados como indicativos de evidência da associação da região cromossômica com o fenótipo investigado. Realizou-se a prospecção de genes candidatos nas regiões detectadas por diferentes metodologias (busca de SNP intragênicos, mapeamento de genes potencialmente em LD com SNP e construção de redes de termos funcionais, além de revisão manual para evidência em processo reprodutivo baseado em literatura científica). Grande número de regiões genômicas foram indicadas pelos efeitos de substituição alélica estimados, os quais apresentaram distribuição esperada para uma característica poligênica. Os resultados obtidos revelaram funções biológicas relacionadas a processos reprodutivos e sugeriram os genes PRKCD (protein kinase C delta type), PRKCE (protein kinase C epsilon type), ITHI-1 (inter-alpha-trypsin inhibitor heavy chain H1 precursor), ITHI-3 (inter-alphatrypsin inhibitor heavy chain H3), ITHI-4 (inter-alpha-trypsin inhibitor heavy chain H4 precursor), BT.62377 (sarcoplasmic/endoplasmic reticulum calcium... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Genome-wide scans for age at first calving (AFC) and gestation length (GL) in Nellore cattle were performed using a panel of over 777,000 single nucleotide polymorphisms (SNP). Association analyses were carried out in three distinct sample groups and their respective phenotypic data: 1) AFC, in months, for 96 cows, 2) AFC and GL estimated breeding values (EBVs) for 831 sires, and 3) AFC and GL EBVs for 1,278 cows. For each dataset and each marker, allele substitution effects were estimated via ordinary least squares and corrected for population stratification. Markers exhibiting high estimated effects were considered as indicative of evidence of association between the chromosomal region and phenotype investigated. Different methodologies were applied for the prospection of candidate genes within the detected regions (intragenic SNP mining, mapping of genes potentially in linkage disequilibrium with SNP, networking of functional terms, and manual review of genes previously found to be involved in reproductive processes). Large amounts of genomic regions were indicated by the estimated allele substitution effects, which presented a distribution consistent with a polygenic trait. The results obtained revealed biological functions related to reproductive processes and suggested PRKCD (protein kinase C delta type), PRKCE (protein kinase C epsilon type), ITHI-1 (inter-alpha-trypsin inhibitor heavy chain H1 precursor), ITHI-3 (interalpha- trypsin inhibitor heavy chain H3), ITHI-4 (inter-alpha-trypsin inhibitor heavy chain H4 precursor), BT.62377 (sarcoplasmic/endoplasmic reticulum calcium ATPase 3), EFCAB6 (uncharacterized gene), TPT1 (translationally-controlled tumor protein) and AMH (muellerian-inhibiting factor precursor), and ZBTB16 (zinc finger and BTB domain-containing protein 16) and PIK3R1 (Phosphatidylinositol 3-kinase... (Complete abstract click electronic access below) / Doutor

Orientador: Roberto Carvalheiro / Coorientador: Henrique Nunes de Oliveira / Banca: Lúcia Galvão de Albuquerque / Banca: Rúsbel Raúl Aspilcueta Borquis / Resumo: A seleção genômica tem sido apontada como uma tecnologia que propiciará aumentos expressivos nas taxas de progresso genético em programas de melhoramento animal. Um dos fatores limitantes para a aplicação da seleção genômica é o custo associado à necessidade de genotipagem de um número grande de animais, com painéis de alta densidade, para a obtenção de boa habilidade de predição dos valores genéticos. Uma alternativa para reduzir custos seria genotipar parte dos animais com painéis menos densos e utilizar técnicas de imputação de genótipos. Em bovinos da raça Nelore, ainda não há consenso quanto à densidade dos painéis a serem utilizados e quanto à estratégia de genotipagem e imputação. Assim, objetivou-se com o presente projeto avaliar a habilidade de predição da seleção genômica utilizando painéis de marcadores moleculares de diferentes densidades, assim como o efeito da utilização de genótipos imputados na habilidade de predição da seleção genômica, em bovinos da raça Nelore. A característica considerada foi precocidade de terminação. Um total de 2035 animais geneticamente avaliados para essa característica e genotipados com o chip Ilumina ® HD Bovina (780k) foram utilizados nas análises, simulando uma situação em que os animais teriam sido genotipados com chips de densidade mais baixa. Análises de seleção genômica também foram executadas utilizando parte dos marcadores do painel de 780k , disponibilizando apenas os SNPs em comum com as seguintes painéis: Illumina® BovineLD (7K), Illumina® BovineSNP50 v2 (50K) e GeneSeek® Genomic Profiler 20K e 75K para Bos indicus. A imputação dos genótipos foi feita com o uso do programa FImpute e as predições genômicas foram conduzidas utilizando os métodos GBLUP e LASSO Bayesiano. Ambas análises, de imputação e de predição genômica, foram conduzidas de forma repetida, seguindo um esquema de validação cruzada, com a formação ... / Abstract: Genomic selection has been considered as a technique that will allow significant increase in the genetic progress of animal breeding programs. A constrain for genomic selection application is the cost of genotyping several animals with high density chips in order to obtain a good prediction equation. An alternative to reduce costs is to genotype part of the animals with a lower density chip and to impute the unobserved genotypes. In Nelore cattle, there is no consensus about which chip and genotyping strategy should be adopted. Therefore, the aim of the present study was to evaluate the predictive ability of genomic selection in Nelore cattle using chips with different densities, and also to assess the effect of using imputed genotypes. The trait considered was finishing precocity. A total of 2,035 animals genetically evaluated for this trait and genotyped with the Illumina® Bovine HD chip (780K) were used in the analyses, mimicking a situation where the animals would have been genotyped with lower density chips. Genomic selection analyses were also run masking part of the 780K genotypes, making available just the SNPs in common with the following chips: Illumina® BovineLD (7K), Illumina® BovineSNP50 v2 (50K), GeneSeek® Genomic Profiler 20K and 75K for Bos indicus. Genomic prediction analyses were run with or without imputing the masked genotypes, using the software FImpute, and under the GBLUP and Bayesian LASSO methods. A 5- fold cross-validation scheme was adopted to perform the analyses, randomly assigning the groups. Results showed that the 50K and 75K chips presented the same predictive ability as the 780K chip. The results also indicated that if the 780K was considered as the target chip for applying genomic selection in the Nelore breed, its cost effectiveness could be improved with the strategy of genotyping part of the animals with a lower density chip (7K or 20K) and imputing their 780K missing genotypes. Further studies ... / Mestre

Orientador: Fernando Sebastian Baldi Rey / Coorientador: William Koury Filha / Coorientador: Claudio de Ulhoa Magnabosco / Coorientador: Rafael Lara Tonussi / Banca: Danísio Prado Munari / banca: Carina Ubirajara Faria / Resumo: O objetivo do presente estudo foi reunir as características morfológicas de estrutura (ES), precocidade (PS) e musculosidade (MS) ao sobreano em características únicas indicadoras de biotipo animal, estimar seus parâmetros genéticos e para características de importância econômica como peso ajustado aos 450 dias de idade (P450), área de olho de lombo (AOL), espessura de gordura subcutânea na costela (EG), espessura de gordura subcutânea na garupa (EGP8), ES, PS, MS, perímetro escrotal ajustado aos 365 (PE365) e 450 (PE450) dias de idade, idade ao primeiro parto (IPP), probabilidade de parto precoce (PPP) e stayability(STAY), a fim de verificar a possibilidade da utilização das características indicadoras biotipo em programas de melhoramento genético de bovinos de corte da raça Nelore. Os dados analisados foram fornecidos pela Associação Nacional de Criadores e Pesquisadores (ANCP). O modelo utilizado para a estimativa de parâmetros genéticos incluiu o efeito aleatório genético aditivo direto e o efeito fixo de grupo de contemporâneos. Para as análises foi utilizada a metodologia ssGBLUP. Métodos multivariados foram empregados para análises de componentes principais pelo procedimento PROC PRINCOMP do software SAS. Foram utilizados dados de ES, PS, MS e P450. O primeiro e o segundo componentes da análise explicaram 82,08% da variabilidade dos dados de e foram utilizados como autovalores para obtenção dos índices de biotipo animal PCA1 e PCA2. A característica SAM foi formada pel... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: The aim of this study was to gather the morphological characteristics after yearling of body structure (BS), finishing precocity (FP) and muscling (MS) in unique traits indicative of animal biotype and estimate its genetic parameters and for the characteristics of economic importance of body weight at 450 days of age (W450), ribeye area (REA), backfat thickness (BF), rump fat thickness (RF), BS, FP, MS, scrotal circumference at 365 (SC365) and 450 (SC450) days of age, age at first calving (AFC), heifer pregnancy (HP) and stayability (STAY) in order to verify the possibility of using the animal biotype indicator characteristics in breeding programs for Nelore beef cattle. The data analyzed were provided by the National Association of Breeders and Researchers (ANCP). The model used for the estimation of genetic parameters included the additive genetic direct random effect and the fixed group effect of contemporaries. For the analyzes, the ssGBLUP methodology was used. Multivariate methods were used for principal component analyzes by the PROC PROCOMP procedure of the SAS software. Data from BS, FP, MS and W450 were used. The first and second components of the analysis explained 82.08% of the variability of the data and were used as eigenvalues to obtain the indices of animal biotype PCA1 and PCA2. The SAM characteristic was formed by the combination of BS, FP and MS scores, penalizing or subsidizing late or early animals, respectively. The animal biotype indexes presented high ... (Complete abstract click electronic access below) / Mestre

La complexe protéique Sir (Silent Information Regulator) de la levure bourgeonnante est l’acteur principal dans la formation de l’hétérochromatine, qui provoque l’atténuation de l’expression génique par un mécanisme épigénétique. Le complexe Sir lié à la chromatine maternelle est démonté lors de la réplication génomique et puis réformé sur les deux brins nouvellement répliqué. La dynamique de maintien de Sir sur la chromatine pendant le cycle cellulaire et dans de variables conditions de croissance n’est pas bien comprise. Pour comprendre comment la structure chromatinienne telle que l’hétérochromatine peut être héritée et par conséquent comment les structures épigénétiques sont transmises d’une génération cellulaire à l’autre, nous avons besoin de mesurer la vitesse d’échange des sous-unités du complexe Sir au cours du cycle cellulaire dans différentes conditions de croissance. Nous avons donc utilisé le système RITE qui permet d’échanger deux épitopes attachés à Sir3 (une des sous-unités de Sir) et par la suite mesurer la cinétique de remplacement de Sir3. Nous avons constaté que la Sir3 maternelle est complètement remplacée par la Sir3 nouvellement synthétisées dans les régions télomériques durant le premier cycle cellulaire après la sortie de la phase stationnaire. Nous proposons que cette reprogrammation du complexe hétérochromatique est un mécanisme d’adaptation qui assure l’activation des gènes de réponse au stress par la déstabilisation transitoire de la structure hétérochromatinienne. / The budding yeast SIR complex (Silent Information Regulator) is the principal actor in heterochromatin formation, which causes epigenetically regulated gene silencing phenotypes. The maternal chromatin bound SIR complex is disassembled during replication and then, if heterochromatin is to be restored on both daughter strands, the SIR complex has to be reformed on both strands to pre-replication levels. The dynamics of SIR complex maintenance and re-formation during the cell-cycle and in different growth conditions are however not clear. Understanding exchange rates of SIR subunits during the cell cycle and their distribution pattern to daughter chromatids after replication has important implications for how heterochromatic states may be inherited and therefore how epigenetic states are maintained from one cellular generation to the next. We therefore used the tag switch RITE system to measure genome wide turnover rates of the SIR subunit Sir3 before and after exit from stationary phase and show that maternal Sir3 subunits are completely replaced with newly synthesized Sir3 at subtelomeric regions during the first cell cycle after release from stationary phase. We propose that the observed “reset” of the heterochromatic complex is an adaptive mechanism that ensures the activation of subtelomeric stress response genes by transiently destabilizing heterochromatin structure.

Orientador: Danísio Prado Munari / Coorientador: Ricardo Vieira Ventura / Coorientador: Roberto Carvalheiro / Banca: Fernando Sebastián Baldi Rey / Banca: Arione Augusti Boligon / Resumo: Painéis de marcadores genéticos de alta densidade (HD) possuem forte desequilíbrio de ligação, que permite melhores predições de valores genômicos. Entretanto, genotipar animais com estes painéis apresenta custo elevado, tornando-se uma limitação para a genotipagem de todos os candidatos à seleção. Uma alternativa para a redução desses custos é utilizar imputação de genótipos. A imputação é um método em que marcadores de uma população genotipada com painéis de baixa densidade (LD) são inferidos utilizando informações provenientes de uma população referência genotipada com painéis HD. O objetivo deste trabalho foi comparar em diferentes cenários metodologias de imputação de marcadores moleculares de polimorfismos de nucleotídeos únicos (SNP) em bovinos de corte da raça Canchim. Foram utilizadas informações de 285 animais da raça Canchim, 114 do grupo genético "MA" e 1 touro da raça Charolês genotipados com painel Illumina BovineHD BeadChip (786.799 SNP), nascidos entre 1999 e 2005 e provenientes da base de dados genômicos da Embrapa Pecuária Sudeste, São Carlos, SP. A edição dos dados foi realizada no software R e em linguagem C++. Para a frequência mínima de alelos (MAF) foram aplicados 3 diferentes critérios: sem remover MAF (QC1); SNP com MAF menor que 0,0025 (QC2) e menor que 0,10 (QC3) foram excluídos. O painel HD original foi reduzido para painéis de baixa densidade (LD) 3K, 6K, 9K, 50K, 20K, 80K e 90K, selecionando os marcadores em comum entre o painel HD original e os painéis comerciais Illumina Bovine3K (3K), BovineLD (6K), GeneSeek Genomic Profiler (GGP) Beef LD (9K), BovineSNP50 (50K), GGP Indicus LD (20K) ,GGP Beef HD (80K) e GGP Indicus HD (90K). Os animais foram divididos em diferentes cenários, denominados de população referência e imputação, sendo o cenário 1 (C1): População referência formada por animais nascidos ... / Abstract: High-density panels (HD) have strong level linkage disequilibrium among genetic markers (i.e. single nucleotide polymorphism - SNP), which allows better predictions of genomic breeding values. However, HD genotyping still expensive and became a limitation for the quantity of candidate animals used in genomic studies. As an alternative to decrease costs, imputation methods are powerful tools to infer missing marker genotypes from low-density (LD) panels to HD. Imputation uses information from a reference population of animals genotyped with a HD panel to impute variants that are not directly genotyped in LD panels. The objective of this study was to compare different scenarios and methodologies of imputation for the Canchim cattle. Data set was provided by Embrapa Pecuária Sudeste and comprised 285 Canchim animals, 114 MA genetic group animals, and 1 ancestor Charolais bull. Animals born between 1999 and 2005 were genotyped with the Illumina BovineHD panel (786,799 SNP). Data editing was performed in the R software and in C ++ language. Multiple scenarios combining different minor allele frequencies (MAF) thresholds for SNPs were tested: no MAF filter (QC1), and exclusion of SNPs with MAF lower than 0.0025 (QC2) and MAF lower than 0.10 (QC3). LD panels were created by masking SNPs originally present in the HD panel, and then assigning markers into the Illumina Bovine3K (3K), Illumina BovineLD (6K), Beef LD GeneSeek Genomic Profiler (9K), Indicus LD GeneSeek Genomic Profiler (20K), Illumina BovineSNP50 (50K), GeneSeek Genomic Profiler Beef HD (80K) and GeneSeek Genomic Profiler Indicus HD (90K) panels. Reference and target populations were defined as scenario 1 (C1), reference animals were born up until 2004 and target animals were born in 2005; scenario 2A (C2A), reference animals from Canchim breed and target animals from MA genetic group; scenario 2B (C2B), reference animals from MA genetic ... / Mestre

Orientador: Danísio Prado Munari / Coorientador: Mônica Correa Ledur / Coorientador: Ricardo Vieira Ventura / Banca: Ricardo da Fonseca / Banca: Mauricio de Alvarenga Mudado Mudadu / Resumo: O objetivo do presente trabalho foi avaliar a acurácia de predição dos valores genéticos genômicos para características de diferentes herdabilidades, em populações simuladas de aves com diferentes níveis de endogamia. Os dados fenotípicos e genotípicos foram simulados com base na estrutura populacional de uma população experimental de aves poedeiras. Foram simulados os fenótipos e os genótipos de aves para características de taxa de postura total de ovos (PTO) e peso dos ovos as 32 semanas de idade (PO), com herdabilidades de 0,15 e 0,37 respectivamente. Foi simulada uma população histórica a fim de gerar desequilíbrio de ligação na população e esta deu origem as populações recentes em que foram simulados três cenários populacionais, visando maximizar (REC1), minimizar (REC2) e aleatorizar (REC3) os acasalamentos endogâmicos. Ao longo de 10 gerações recentes, os animais foram selecionados com base nos maiores valores genéticos preditos (VGP), utilizando o BLUP (best linear unbiased prediction) tradicional. O genoma das aves foi simulado ao longo dos 958 Mb do genoma Gallus gallus 4.0 com 3.747 QTL (loci de caracteres quantitativos) aleatoriamente distribuídos e 49.978 marcadores SNP uniformemente distribuídos. A fim de alterar as frequências alélicas e gerar variabilidade genética ao longo das gerações, foram simulados eventos de deriva genética, taxa de recombinação e mutação recorrente. Para avaliar os efeitos da endogamia nas populações recentes, foram calculados os desequilíbrios de ligação (DL), o tamanho efetivo da população (Ne) e as tendências genéticas em todos os cenários de populações recentes em ambas as características simuladas. Para predizer os valores genéticos genômicos preditos (VGGP), as populações recentes foram subdivididas em populações de treinamento e validação. Nas subpopulações de treinamento, os 960 animais ... / Abstract: The objective of this study was to evaluate the prediction accuracy of genomic breeding values for traits of different heritability in simulated populations with different inbreeding. Phenotypic and genotypic data were simulated based on the population structure of an experimental population of White Leghorn hens at Embrapa Suínos e Aves. The phenotypes and genotypes were simulated for the rate of total egg production (PTO) and egg weight to 32 weeks of age (PO) with heritability of 0.15 and 0.37, respectively. The historical population was simulated to generate linkage disequilibrium in the population. Three scenarios in recent populations were simulated for each trait: REC1, REC2 and REC3 to maximize inbreeding, minimize inbreeding and random mating, respectively. The animals were selected based on the largest breeding values along 10 generations. The genome of the birds was simulated with eight macro-chromosomes and 19 micro-chromosomes with 3.747 QTL randomly distributed and 49.978 SNPs markers evenly spaced along the 958 cM. Recombination, random drift and recurrent mutation were simulated in order to generate genetic variability. The linkage disequilibrium (LD), effective population (Ne) and genetic trends were calculated for all scenarios. Each recent population was divided in training and validation sets In order to predict the genomic breeding values. The training set included the genotypes and phenotypes of 960 animals, which had higher breeding values' accuracy. The validation set had 1120 animals of the last generation of the recent population. The average inbreeding ranged from 0.06 ± 0.30 to 0.22 ± 0.12 for PTO and 0.05 ± 0.03 to 0.20 ± 0.12 for PO. The REC1 populations had higher inbreeding along generations compared, both for PTO and PO, compared to REC2 and REC 3, and consequently higher level of LD. The highest accuracy for PTO and PO were ... / Mestre