Precision medicine provides targeted treatment for an individual patient based on disease mechanisms, promoting health care. Matched transcriptomes derived from a single subject enable uncovering patient-specific dynamic changes associated with disease status. Yet, conventional statistical methodologies remain largely unavailable for single-subject transcriptome analysis due to the "single-observation" challenge. We hypothesize that, with statistical learning approaches and large-scale inferences, one can learn useful information from single-subject transcriptome data by identifying differentially expressed genes (DEG) / pathways (DEP) between two transcriptomes of an individual. This dissertation is an ensemble of my research work in single-subject transcriptome analytics, including three projects with varying focuses. The first project describes a two-step approach to identify DEPs by employing a parametric Gaussian mixture model followed by Fisher's exact tests. The second project relaxes the parametric assumption and develops a nonparametric algorithm based on k-means, which is more flexible and robust. The third project proposes a novel variance stabilizing framework to transform raw gene counts before identifying DEGs, and the transformation strategically by-passes the challenge of variance estimation in single-subject transcriptome analysis. In this dissertation, I present the main statistical methods and computational algorithms for all the three projects, as well as their real-data applications to personalized treatments.

Doctor of Philosophy / Department of Plant Pathology / Christopher R. Little / Charcoal rot, caused by the necrotrophic fungus, Macrophomina phaseolina (Tassi) Goid., is an important disease in sorghum (Sorghum bicolor (L.) Moench). The molecular interactions between sorghum and M. phaseolina are poorly understood. In this study, a large-scale RNA-Seq experiment and four follow-up functional experiments were conducted to understand the molecular basis of charcoal rot resistance and/or susceptibility in sorghum. In the first experiment, stalk mRNA was extracted from charcoal-rot-resistant (SC599) and susceptible (Tx7000) genotypes and subjected to RNA sequencing. Upon M. phaseolina inoculation, 8560 genes were differentially expressed between the two genotypes, out of which 2053 were components of 200 known metabolic pathways. Many of these pathways were significantly up-regulated in the susceptible genotype and are thought to contribute to enhanced pathogen nutrition and virulence, impeded host basal immunity, and reactive oxygen (ROS) and nitrogen species (RNS)-mediated host cell death. The paradoxical hormonal regulation observed in pathogen-inoculated Tx7000 was characterized by strongly upregulated salicylic acid and down-regulated jasmonic acid pathways. These findings provided useful insights into induced host susceptibility in response to this necrotrophic fungus at the whole-genome scale. The second experiment was conducted to investigate the dynamics of host oxidative stress under pathogen infection. Results showed M. phaseolina’s ability to significantly increase the ROS and RNS content of two charcoal-rot-susceptible genotypes, Tx7000 and BTx3042. Over-accumulation of nitric oxide (NO) in stalk tissues in the pathogen-inoculated susceptible genotypes was confirmed using a NO-specific fluorescent probe and confocal microscopy. Significantly increased malondialdehyde content confirmed the enhanced oxidative stress experienced by the susceptible genotypes after pathogen inoculation. These findings suggested the contribution of oxidative stress-associated induced cell death on charcoal rot susceptibility under infection. In the third functional experiment, the behavior of the sorghum antioxidant system after pathogen inoculation was investigated. M. phaseolina significantly increased the glutathione s-transferase (GST), glutathione peroxidase (GPX), glutathione reductase (GR), and peroxidase activities of the susceptible genotypes (Tx7000, BTx3042) but not in the resistant genotypes (SC599, SC35). Increased activities of these enzymes in susceptible genotypes may contribute to reduced oxidative stress thus lowering charcoal rot susceptibility. The fourth functional experiment was designed to quantify induced host-derived cell wall degrading enzymes (CWDEs) using crude enzyme mixtures from stalks. A gel diffusion assay revealed significantly increased pectinesterase activity in pathogen-inoculated Tx7000 and BTx3042 while significantly increased polygalacturonase activity was determined by absorbance. Fluorimetric determination of cell extracts revealed significantly increased cellulose degrading enzyme activity in M. phaseolina-inoculated Tx7000 and BTx3042. These findings revealed the pathogen’s ability to promote charcoal rot susceptibility in grain sorghum through induced host CWDEs. The last functional study was designed to profile the stalk tissue lipidome of Tx7000 and SC599 after M. phaseolina inoculation using automated direct infusion electrospray ionization-triple quadrupole mass spectrometry (ESI-MS/MS). M. phaseolina significantly decreased the phytosterol, phosphatidylserine, and ox-lipid contents in Tx7000 while significantly increasing stigmasterol:sitosterol ratio. Except for ox-lipid content, none of the above was significantly affected in resistant SC599. Results suggested the lethal impacts of M. phaseolina inoculation on plastid- and cell- membrane integrity and the lipid-based signaling capacity of Tx7000. Findings shed light on the host lipid classes that contribute to induced charcoal rot susceptibility in grain sorghum.

Le développement de la seconde génération de séquenceurs haut débit a généralisé l'accès à l'étude du transcriptome via le protocole RNAseq. Celui-ci permet d'obtenir à la fois la séquence et l'abondance des transcrits d'un échantillon. De nombreuses méthodes bioinformatiques ont été et sont encore développées pour permettre l'analyse des données issues du RNAseq et en tirer le maximum d'information. Ce type d'analyse est notamment possible sans utiliser de génome de référence, et donc pour les espèces modèles ou non-modèles, grâce à des méthodes d'assemblage. Durant ma thèse, j'ai principalement travaillé à partir de données RNA-seq issues d'espèces non modèles. Je me suis intéressée dans un premier temps à l'impacte de l'hybridation inter spécifique sur la stabilité des génomes chez les hybrides issus des croisements réciproques de D. mojavensis et D. arizonae. Nos résultats ne montrent pas une dérégulation globale, mais plutôt quelques gènes et éléments transposables qui sont spécifiquement dérégulés. La pipeline d'analyse mis en place ici sera réutilisée pour l'étude des niveaux d'expression des transcrits chez les mâles ainsi que pour les croisements issus d'autres lignées de D. mojavensis avec D. arizonae, conduisant à une fertilité variable chez les hybrides.Dans un second temps, j'ai participé à la validation du logiciel KisSplice pour la détection de SNP dans des données RNA-seq sans génome de référence. Celui-ci permet de trouver différents types de variants (épissage, indels) directement dans le graphe de de Bruijn construit à partir des lectures séquencées. J'ai également participé au développement d'outils de post-traitement permettant de prédire l'impact des SNP sur les protéines / Next-generation high throughput sequencing technologies provide efficient, rapid, and low cost access to sequencing. Its application to transcriptomes, called RNA-seq, enables the study of both the sequence and the expression of the transcripts. Many bio-informatics methods are still developed for RNA-seq data processing, trying to get the maximum out of it. Assembly methods allow us to study non-model species (no reference genome available) as well as model species. The work presented here is mostly related to RNA-seq data on non-model species.In the first study, to understand the initiation of hybrid incompatibility, we performed a genome-wide transcriptomic analysis on ovaries from parental lines and on hybrids from reciprocal crosses of \emph{D. mojavensis} and \emph{D. arizonae}. We didn't see a global deregerulation of genes or transposable element. Instead, we show that reciprocal hybrids presented specific gene categories and few transposable element families misexpressed relative to the parental lines. The analytical workflow developed for this project will be used to analyze transcriptomic data from the testis, but also to study the reciprocal crosses from other lines of D. mojavensis with D. arizonae leading to variable levels of sterility in hybrids. A second project tacked here is the identification and quantification of SNPs from RNA-seq data without a reference genome with KisSplice. Kissplice was developed to identified several type of variants (splicing events, indels) directly from the de Bruijn graph, build from the sequenced reads. We also developed other KisSplice-tools, for downstream analyses of the SNPs, including the prediction of their impact on the protein sequence

We study the tasks of transcript expression quantification and differential expression analysis based on data from high-throughput sequencing of the transcriptome (RNA-seq). In an RNA-seq experiment subsequences of nucleotides are sampled from a transcriptome specimen, producing millions of short reads. The reads can be mapped to a reference to determine the set of transcripts from which they were sequenced. We can measure the expression of transcripts in the specimen by determining the amount of reads that were sequenced from individual transcripts. In this thesis we propose a new probabilistic method for inferring the expression of transcripts from RNA-seq data. We use a generative model of the data that can account for read errors, fragment length distribution and non-uniform distribution of reads along transcripts. We apply the Bayesian inference approach, using the Gibbs sampling algorithm to sample from the posterior distribution of transcript expression. Producing the full distribution enables assessment of the uncertainty of the estimated expression levels. We also investigate the use of alternative inference techniques for the transcript expression quantification. We apply a collapsed Variational Bayes algorithm which can provide accurate estimates of mean expression faster than the Gibbs sampling algorithm. Building on the results from transcript expression quantification, we present a new method for the differential expression analysis. Our approach utilizes the full posterior distribution of expression from multiple replicates in order to detect significant changes in abundance between different conditions. The method can be applied to differential expression analysis of both genes and transcripts. We use the newly proposed methods to analyse real RNA-seq data and provide evaluation of their accuracy using synthetic datasets. We demonstrate the advantages of our approach in comparisons with existing alternative approaches for expression quantification and differential expression analysis. The methods are implemented in the BitSeq package, which is freely distributed under an open-source license. Our methods can be accessed and used by other researchers for RNA-seq data analysis.

Orientadores: Felipe Rodrigues da Silva, Francisco Pereira Lobo / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Biologia / Made available in DSpace on 2018-08-27T12:42:35Z (GMT). No. of bitstreams: 1 Castro_GiovanniMarquesde_M.pdf: 8706589 bytes, checksum: bd5b5d6428a549bafa82854367f811b9 (MD5) Previous issue date: 2015 / Resumo: A planta Fevillea cordifolia L. possui um grande potencial para produção de biodiesel. Buscando entender o metabolismo foi realizado um experimento exploratório de RNA-seq com sementes inteiras. No entanto, as análises da qualidade na biblioteca indicaram grande quantidade de sequências virais. Após a reconstrução do transcriptoma usando o programa TRINITY, também foi reconstruído o genoma completo do vírus. O vírus reconstruído possui identidade de 96% com o Papaya ringspot virus (PRSV) em um alinhamento global de nucleotídeos dos genomas completos. Avaliando a abundância dos contigs, o PRSV encontrado representa quase 60% das 24,6 milhões de leituras da biblioteca. Para identificar qual a origem do vírus encontrado, este foi comparado com 29 PRSVs existentes no Genbank através de análise filogenética usando do Algerian watermelon mosaic virus (AWMV) para enraizar a árvore. O vírus encontrado agrupa-se com os dois PRSVs do Brasil, dentro do grupo das Américas estando mais próximo do PRSV-W-C (DQ374152). A existência de recombinações entre os PRSVs foi analisada, porém não foi detectada recombinação recente. Devido à profundidade de sequenciamento maior que 10.000x, foi possível analisar as variações existentes do genoma viral reconstruído. Uma análise das variações dos códons virais foi realizada, mostrando uma tendência para ocorrência de indels para a região do genoma no fim do cístron NIb e início do cístron CP. Essas variações ocorrem devido à existência de haplótipos virais na amostra sequenciada. Para se estimar a diversidade de haplótipos virais da amostra, foi realizada uma reconstrução local da região de 500 nt com mais alta entropia. Foram reconstruídos 58 haplótipos, dos quais dois eram predominantes com frequências de 64,84% e 24,34%. Os haplótipos reconstruídos podem ser usados para o desenvolvimento de resistência mediada por RNAi, evitando que uma variante conhecida e preexistente na população possa quebrar a resistência / Abstract: The plant Fevillea cordifolia L. has a great potential for producing biodiesel. In order to understand its metabolic pathways an exploratory RNA-seq experiment was conducted with its whole-seeds. However the quality analysis of the library revealed a substantial amount of virus sequences. After the reconstruction of the transcriptome using the software TRINITY, the complete viral genome was obtained as well. The reconstructed viral genome had an identity of 96% with Papaya ringspot virus (PRSV) in a global nucleotide alignment using whole genomes. When estimating the abundance of the reconstructed sequences, this PRSV had almost 60% of the 24,6 million reads mapping to it. Aiming to elucidate the origin of this virus, its sequence was compared to 29 PRSVs from Genbank using phylogenetic analysis and the Algerian Watermelon Mosaic Virus (AWMV) as an outgroup. This PRSV clustered with the Brazilian isolates, being closer to PRSV-W-C (DQ374152). A recombination analysis was performed within the PRSVs but no recent recombination was detected. Due to the depth of the coverage sequencing being higher than 10.000x, it was possible to analyze the variations existing in the reconstructed genome. An analysis of the codons variations was performed, revealing a tendency for the occurrence of indels in a region at the end of NIb cistron and at the start of the CP cistron. These variations occur due to the existence of viral haplotypes sequenced in this sample. A local reconstruction of the 500nt region with the highest entropy was performed to estimate the diversity of viral haplotypes in this sample. 58 haplotypes were reconstructed, of which 2 were dominant with frequencies of 64,84% and 24,34%. The reconstructed haplotypes may be used for the development of RNAi-mediated resistance, avoiding the breaking of the resistance by variants that are known to exist in the population / Mestrado / Bioinformatica / Mestre em Genética e Biologia Molecular

No description available.

L’analyse des défauts d’épissage est particulièrement complexe. Outre la diversité des transcrits présents à l’état physiologique, les variations nucléotidiques peuvent induire des modifications hétéroclites de l’épissage. Ces variations, appelées variants splicéogéniques, et leur impact au niveau de l’épissage, sont à même de modifier plus ou moins sévèrement le phénotype de l’individu.Au cours de ce travail de thèse, nous nous sommes intéressés à trois grands aspects de l’étude des défauts de l’épissage : (i) la prédiction de ces défauts d’épissage, (ii) l’analyse des données de RNA-seq et (iii) le rôle de l’épissage dans l’interprétation de la pathogénicité d’un variant pour la prédisposition aux cancers du sein et de l’ovaire (syndrome HBOC).Nous avons optimisé les recommandations en vigueur pour identifier les variants splicéogéniques au sein des séquences consensus des sites d’épissage. Ce travail a conduit à la publication d’un nouvel outil SPiCE (Splicing Prediction in Consensus Elements), développé sur 395 variants. SPiCE a le potentiel d’être une aide à la décision pour guider les généticiens vers ces variants splicéogéniques, grâce à une exactitude de 94.4 %. Puis, nous avons comparé les outils de prédiction des points de branchement. Pour cela, une collection sans précédente de 120 variants avec leurs études ARN a été établi dans la région des points de branchements. Nous avons ainsi révélé que ces outils de prédictions sont aptes à prioriser les variants pour des études ARN dans ces régions jusque-là peu étudiées. Pour étendre les prédictions des variants splicéogéniques au-delà d’un motif spécifique, nous avons construit l’outil SPiP (Splicing Prediction Pipeline). SPiP utilise un ensemble d’outils pour prédire un défaut d’épissage quel que soit la position du variant. Ainsi, SPiP peut ainsi s’adresser à la diversité des défauts d’épissage avec une exactitude de 80.21 %, sur une collection de 2 784 variants.Les données issues du RNA-seq sont complexes à analyser, car il existe peu d’outils pour annoter finement les épissages alternatifs. Aussi nous avons publié l’outil SpliceLauncher. Cet outil permet de déterminer une grande diversité de jonctions d’épissage, indépendamment des systèmes RNA-seq utilisés. Cet outil renvoie aussi les résultats sous formes graphiques pour faciliter leur interprétation.Puis nous avons évalué le rôle de l’épissage alternative dans l’interprétation à usage clinique d’un variant. Le gène PALB2, impliqué dans le syndrome HBOC, a été utilisé comme modèle d’étude. Nous avons ainsi démontré que l’épissage alternatif de PALB2 est apte à remettre en cause la pathogénicité de certains variants. La collecte de données fonctionnelles et cliniques sont donc nécessaires pour conclure sur leur pathogénicité.Nos travaux illustrent ainsi l’importance de la caractérisation et de l’interprétation des modifications de l’épissage pour répondre aux défis présents et futurs du diagnostic moléculaire en génétique. / Analysis of splicing defects is particularly complex. In addition to the diversity of physiological transcripts, nucleotidic variations can induce heterogeneous alteration of splicing. These variations, called spliceogenic variants, and their impact on splicing, can involve severe consequences on the individual phenotype.In this thesis work, we focused on three main aspects of the study of splicing defects: (i) the prediction of these splicing defects, (ii) the analysis of RNA-seq data and (iii) the role of splicing in interpreting the pathogenicity of a variant for the hereditary breast and ovarian cancers (HBOC syndrome).We optimized the current recommendations to identify spliceogenic variants within the consensus sequences of splicing sites. This work led to the publication of a new tool, SPiCE (Splicing Prediction in Consensus Elements), developed on 395 variants. SPiCE has the potential to be a decision support tool to guide geneticists towards these spliceogenic variants, with an accuracy of 94.4%. Then, we compared the tools dedicated to branch points prediction. For this purpose, an unprecedented collection of 120 variants with their RNA studies has been established in the branch point region. Thus, we revealed these prediction tools are able to prioritize variants for RNA studies in these hitherto poorly studied regions. To extend the predictions of spliceogenic variants beyond a specific motif, we built SPiP (Splicing Prediction Pipeline) tool. SPiP uses a set of tools to predict a splicing defect regardless of the variant position. Thus, SPiP can address the diversity of splicing defects with an accuracy of 80.21%, on a collection of 2,784 variants.The data from the RNA-seq are complex to analyze, as there are few tools to finely annotate alternative splices. Also we published SpliceLauncher tool. This tool allows to determine a wide variety of splicing junctions, independently of RNA-seq systems used. This tool also returns the results in graphical form to make interpretation user-friendly.Then we evaluated the role of alternative splicing in the clinical interpretation of a variant. The PALB2 gene, involved in HBOC syndrome, was used as a study model. Thus, we demonstrated that the alternative splicing of PALB2 is able of challenging the pathogenicity of certain variants. Collection of functional and clinical data is therefore necessary to conclude on their pathogenicity.Our work thus illustrates the importance of characterizing and interpreting splicing modifications to meet the current and future challenges of molecular diagnosis in human genetics.

Urinary tract infections (UTI) are among the most common infections in humans. Uropathogenic Escherichia coli (UPEC), the main causative agent of UTIs, can invade and replicate within bladder epithelial cells, and recent evidence demonstrated that some UPEC strains also survive within macrophages. To understand the mechanisms of host subversion that enable UPEC to survive within macrophages, and the contribution of macrophages to UPEC-mediated pathology, I performed host-pathogen co-transcriptome analyses using RNA sequencing. I developed an effective computational framework that simultaneously separated, annotated, and quantified the mammalian and bacterial transcriptomes. First, mouse bone morrow-derived macrophages (BMM) were challenged over a 24 h time course with UPEC reference strains, UTI89 (cystitis strain), 83972 and VR50 (asymptomatic bacteriuria strains) that possess contrasting intramacrophage phenotypes. My results showed that BMM responded to the three different UPEC strains with broadly similar gene expression programs. In contrast to the conserved pattern of BMM responses, the transcriptional responses of the different UPEC strains diverged markedly from each other. Hypothesizing that genes upregulated at 24 h post-infection may contribute to intramacrophage survival, I identified UTI89 genes upregulated at this time point, and showed that deletion of one of these genes (pspA) compromised intramacrophage survival of UPEC strain UTI89. Second, human monocyte-derived macrophages (HMDM) and BMM were challenged over a 24 h course with the UPEC strain EC958, a globally disseminated, multi-drug resistant strain. My analysis identified extensive divergence in UPEC-regulated orthologous gene expression between HMDM and BMM, and I validated both known and novel genes in the context of differential regulation. On the contrary, the transcriptional response of EC958 showed a broad conservation across both mammalian intramacrophage environments. My study thus provides both a unique co-culture approach to study infection in vitro and a technological framework for simultaneously capturing global changes in host-pathogen interactions at the transcriptional level in co-cultures. In conclusion, this work has generated new insights into the mechanisms that UPEC strains exploit to persist within the mouse intramacrophage environment, as well as differences in the transcriptional repertoire of HMDM and BMM challenged with the same UPEC strain.

Diplomová práce se zabývá studiem genové regulace v Clostridium beijerinckii NRRL B-598, pro následné odvození genové regulační sítě bakterie C. beijerinckii NRRL B-598. V teoretické části této práce je uvedena obecná nomenklatura problematiky genové regulace se zaměřením na nomenklaturu genových regulačních sítí. Následně jsou zde popsané laboratorní metody, sloužící pro získání vhodných dat popisující expresi genů. Tato data jsou základem pro studium genové regulace a návrhy genových regulačních sítí. Práce se zaměřuje především na technologii RNA-Seq a stručný popis laboratorních dat získaných ze zmíněné bakterie C. beijerinckii NRRL B-598. V praktické části se práce zabývá předzpracováním těchto surových laboratorních dat a následným studiem genové regulace se zaměřením na odvození operonů a vytvoření prvních genových regulačních sítí pomocí různých přístupů pro C. beijerinckii NRRL B-598.

No description available.