A sequência de 35 aminoácidos da alça V3 da gp120 do gene env do HIV-1 é o principal determinante do tropismo viral pelos correceptores CCR5 ou CXCR4, utilizados pelo HIV-1 para a entrada na célula. O desenvolvimento de estratégias antirretrovirais baseadas no uso dos correceptores representa um avanço importante para o controle da progressão da infecção. Entretando, o uso clínico dos antagonistas de CCR5 implica na determinação do tropismo das cepas virais do indivíduo infectado e os programas preditores de bioinformática para a determinação do tropismo poderiam ser uma alternativa mais acessível para a triagem dos candidatos ao uso dos antagonistas de CCR5. Este estudo teve como objetivo utilizar ferramentas de bioinformática para a predição de tropismo e avaliar sua aplicabilidade na prática clínica. Foram coletadas amostras de sangue periférico de 101 indivíduos infectados pelo HIV-1 e sob acompanhamento clínico, dos quais foram extraídas amostras de DNA proveniente de PBMCs. As amostras de DNA foram amplificadas por PCR para a região da alça V3, das quais foram obtidas 94 sequências. Os sistemas preditivos foram avaliados utilizando 185 sequências com tropismo conhecido provenientes de banco de dados. Com base nesta análise foi possível elaborar um algoritmo para a predição do tropismo com 94% de confiabilidade. Assim, a predição das 94 amostras demonstrou uma prevalência de 80% (n=75) de cepas R5 e 20% (n=19) de cepas X4. Os sistemas preditivos de tropismo podem representar uma importante estratégia para a triagem dos candidatos ao uso dos antagonistas de coreceptor, porém, não são capazes de substituir completamente os ensaios padrão-ouro para a determinação do tropismo. / The 35 amino acids of the V3-gp120 of HIV-1 env gene is the main determinant of viral tropism by the coreceptors CCR5 and CXCR4 used for HIV-1 cell entry. The development of antiretroviral strategies based on the coreceptor usage represents an important step to control the infection progression. However, the clinical application of CCR5 antagonists involves the coreceptor usage determination of viral strains in the infected individual. The bioinformatics predictive programs for coreceptor usage determination could be a more available alternative for screening candidates to receive CCR5 antagonists. This study aimed to employ bioinformatics tools to predict tropism and assess its applicability in clinical practice. Peripheral blood samples were collected from 101 individuals infected with HIV-1 and under clinical follow-up. DNA samples were extracted from PBMCs. The DNA samples were amplified by PCR and 94 V3 sequences were obtained. The predictive systems were evaluated using 185 sequences of known tropism from a database. This analysis provides the construction of an algorithm showing 94% of reliability. Thus, the 94 sample prediction showed a prevalence of 80% (n=75) of R5 strain and 20% (n=19) of X4 strain. The predictive systems could be an important strategy in the screening of the tropism. Nonetheless, they are not able to fully replace the coreceptor usage gold-standard assays.

Alternative splicing of pre-mRNA is a critical mechanism for enabling genetic diversity, and is a carefully regulated process in neuronal differentiation. RNA binding proteins (RBPs) are developmentally expressed and physically interact with RNA to drive specific splicing changes. This work tests the hypothesis that RBP-RNA interactions are critical for regulating timed and coordinated alternative splicing changes during neurodevelopment and that these splicing changes are in turn part of major regulatory mechanisms that underlie morphological and functional maturation of neurons. I describe our efforts to identify functional RBP-RNA interactions, including the identification of previously unobserved splicing events, and explore the combinatorial roles of multiple brain-specific RBPs during development. Using integrative modeling that combines multiple sources of data, we find hundreds of regulated splicing events for each of RBFOX, NOVA, PTBP, and MBNL. In the neurodevelopmental context, we find that the proteins control different sets of exons, with RBFOX, NOVA, and PTBP regulating early splicing changes and MBNL largely regulating later splicing changes. These findings additionally led to the observation that CNS and sensory neurons express a variety of different RBP programs, with many sensory neurons expressing a less mature splicing pattern than CNS neurons. We also establish a foundation for further exploration of neurodevelopmental splicing, by investigating the regulation of previously unobserved splicing events.

Knowledge of the phenotypic states of immune cells in the tumor microenvironment is essential to understand immunological mechanisms of cancer progression, responses to cancer immunotherapy, and the development of novel rational treatments. Yet, this knowledge is opaque to traditional bulk sequencing methods, and novel single-cell RNA sequencing (scRNA-seq) methods which could potentially address these questions introduce complex patterns of error into data that are poorly characterized. This dissertation describes a computational framework, SEQC, built to facilitate rapid and agile analysis of scRNA-seq approaches that utilize molecular barcodes. It combines SEQC with a clustering and normalization method, BISCUIT, and approaches to examine phenotypic diversity and gene variation. These methods are applied to address the unique computational challenges inherent to analysis of single-cell RNA-seq data derived from multiple patients with diverse phenotypes. This dissertation describes an experiment comprising scRNA-seq of over 47,000 immune cells collected from primary breast carcinomas, matched normal breast tissue, peripheral blood, and using these computational approaches. This atlas revealed significant similarity between normal and tumor tissue resident immune cells. However, it also describes continuous tumor-specific phenotypic expansions driven by distinct environmental cues. These results argue against discrete activation states in T cells and the polarization model of macrophage activation in cancer, and have important implications for characterizing tumor-infiltrating immune cells.

Patient-provided data are crucial to achieving the goal of precision medicine. These data, which include family medical history, race and ethnicity, and social and behavioral determinants of health, are essential for disease risk assessment. Despite the well-established importance of patient-provided data, there are many data quality challenges that affect how this information can be used for biomedical research. To determine how to best use patient-provided data to assess disease risk, the research reflected in this dissertation was divided into three overarching aims. In Aim 1, I focused on determining the quality of race and ethnicity, family history and smoking status in clinical databases. In Aim 2, I assessed the impact of various interventions on the quality of these data, including policy changes such as the implementation of the requirements imposed by the Meaningful Use program, and patient-facing tools for collecting and sharing information with patients. In addition to these evaluations, I also developed and evaluated a method “Relationship Inference from the Electronic Health Record” (RIFTEHR), that infers familial relationships from clinical datasets. In Aim 3, I used patient-provided data to assess disease risk both at a population level, by estimating disease heritability, and at an individual level, by identifying high-risk individuals eligible for additional screening for a common disease (diabetes mellitus) and a rare disease (celiac disease). My research uncovered several data quality concerns for patient-provided data in clinical databases. When assessing the impact of interventions on the quality of these data, I found that policy interventions led to more data collection, but not necessarily to better data quality. In contrast, patient-facing tools did increase the quality of the patient-provided data. In the absence of high-quality patient-provided data for family medical history, I developed and evaluated a method for inferring this information from large clinical databases. I demonstrated that electronic health record data can be used to infer familial relationships accurately. Moreover, I showed how the use of clinical data in conjunction with the inferred familial relationships could determine disease risk in two studies. In the first study, I successfully computed disease heritability estimates for 500 conditions, some of which had not been previously studied. In the second study, I identified that screening rates among family members that are considered to be at high-risk for disease development were low for both diabetes mellitus and celiac disease. In summary, the work represented in this dissertation contributes to the understanding of the quality of patient-provided data, how interventions affect the quality of these data, and how novel methods can be applied to troves of existing clinical data to generate new knowledge to support research and clinical care.

Exome sequencing (ES) has empowered genetic diagnosis and novel gene discovery, and is increasingly applied as a first-line test for a variety of disorders. Chronic kidney disease (CKD) affects more than in 1 in 10 persons worldwide, resulting in high morbidity, mortality, and healthcare costs. As CKD displays substantial genetic and phenotypic heterogeneity, the unbiased approach of ES can help to pinpoint a specific etiology and thereby support personalized care. However, the broader utility of ES for nephropathy and challenges associated with such expanded implementation have yet to be systematically assessed. Here, we investigate these questions through integrating ES and phenotype data from large CKD case and control cohorts. First, we survey the genetic and clinical disease spectrum of Mendelian forms of kidney and genitourinary disease, and generate a comprehensive curated list of gene-disease pairs. We then use ES data from 7,974 self-declared healthy adults to evaluate the population prevalence of candidate pathogenic variants for Mendelian nephropathy under different analytic filtering pipelines. We observe an appreciable frequency of putatively diagnostic variants for these conditions using stringent as well as standard filters, resulting in a considerable burden for both variant interpretation and clinical follow-up. Next, we perform ES and diagnostic analysis in a combined cohort of 3,315 all-cause CKD cases. We find diagnostic variants among patients spanning clinical disease categories, and that both the primary and secondary genetic findings resulting from ES have meaningful implications for medical management. We conclude by discussing the greater insights regarding the value of ES for kidney disease emerging from our investigations, and promising avenues for subsequent studies.

No description available.

This work presents the development, analysis and subsequent simulations of mathematical models aimed at providing a basis for modelling atherosclerosis. This cardiovascular disease is characterized by the growth of plaque in artery walls, forming lesions that protrude into the lumen. The rupture of these lesions contributes greatly to the number of cases of stroke and myocardial infarction. These are two of the main causes of death in the UK. Any work to understand the processes by which the disease initiates and progresses has the ultimate aim of limiting the disease through either its prevention or medical treatment and thus contributes a relevant addition to the growing body of research. The literature supports the view that the cause of atherosclerotic lesions is an in inflammatory process-succinctly put, excess amounts of certain biochemical species fed into the artery wall via the bloodstream spur the focal accumulation of extraneous cells. Therefore, suitable components of a mathematical model would include descriptions of the interactions of the various biochemical species and their movement in space and time. The models considered here are in the form of partial differential equations. Specifically, the following models are examined: first, a system of reaction-diffusion equations with coupling between surface and bulk species; second, a problem of optimisation to identify an unknown boundary; and finally, a system of advection-reaction-diffusion equations to model the assembly of keratin networks inside cells. These equations are approximated and solved computationally using the finite element method. The methods and algorithms shown aim to provide more accurate and efficient means to obtain solutions to such equations. Each model in this work is extensible and with elements from each model combined, they have scope to be a platform to give a fuller model of atherosclerosis.

One of evolution's greatest innovations was group living; indeed, it is fundamental to our daily lives as humans. Yet despite intense theoretical and empirical work, the details of how group living arose and is maintained are poorly understood. A central question in this area concerns the strength of natural selection operating between groups of organisms (group selection) because some think this is key to the evolution of group behaviour. It is, however, challenging to measure natural selection occurring between groups and between the individuals within those groups simultaneously. Consequently, a number of contentious theoretical issues have plagued group selection research for a number of decades, and empirical work on this topic is often misinterpreted. In this thesis, I investigate three biological systems that are candidates for group selection where empirical data is readily available. Using techniques from theoretical and computational biology - simulations, game theory and population genetics - I model evolution occurring at multiple levels simultaneously (multi-level selection), shedding light on the evolution and maintenance of group traits. First, I consider the evolution of a trait - lateralization - at the population- and colony-level in eusocial organisms, which have a reproductive structure that promotes group organisation and cooperation. I provide an evolutionary explanation for the strength of lateralization in colonies of the red wood ant, Formica rufa, as a compromise between intraspecific and predatory interactions. After extending the analysis to involve predators targeting multiple colonies simultaneously, I show that populations should tend towards an equal distribution of left- and right- lateralized colonies, resulting in zero population-level lateralization. This contradicts the established view that sociality should produce strong levels of lateralization at the population level. Second, I study a sub-social spider, Anelosimus studiosus, which is a group-living species that has recently been claimed to exhibit group-level adaptation. I use evolutionary game theory to explain the evolution of colony aggression with individual costs and benefits, providing an alternative to the existing group-level interpretation. The model generates a striking fit to the data without any between-group interactions. Therefore, I conclude that more evidence is needed to infer group-level adaptation in this colonial spider. Third, I study the Solanaceae, a plant family whose breeding system is reported to have undergone species selection - group selection acting on whole species. I investigate the evolution of self-fertilization over the family's phylogenetic history. By integrating an existing phylogeny with models of breeding system evolution at the individual level, I find the average selection pressure - and attendant properties of populations - expected to have characterised the Solanaceae over ~36 million years. In conclusion, I have shown the power of modelling approaches to clarify evolutionary explanations, to question existing interpretations, and to identify experiments that can help researchers identify the true causes of trait evolution.

The use of drug combinations (DCs) in cancer therapy can prevent the development of drug resistance and decrease the severity and number of side effects. Synthetic lethality (SL), a genetic interaction wherein two nonessential genes cause cell death when knocked out simultaneously, has been suggested as a method of identifying novel DCs. A combination of two drugs that mimic genetic knockout may cause cellular death through a synthetic lethal pathway. Because SL can be context-specific, it may be possible to find DCs that target SL pairs in tumours while leaving healthy cells unscathed. However, elucidating all synthetic lethal pairs in humans would take more than 200 million experiments in a single biological context – an unmanageably large search space. It is thus necessary to develop computational methods to predict human SL. In this thesis, we develop connectivity homology, a novel measure of network similarity that allows for the comparison of interspecies protein-protein interaction networks. We then use this principle to develop Species-INdependent TRAnslation (SINaTRA), an algorithm that allows us to predict SL between species using protein-protein interaction networks. We validate it by predicting SL in S. pombe from S. cerevisiae, then generate over 100 million SINaTRA scores for putative human SL pairs. We use these data to predict new areas of cancer combination therapy, and then test fifteen of these predictions across several cell lines. Finally, in order to better understand synergy, we develop DAVISS (Data-driven Assessment of Variability In Synergy Scores), a novel way to statistically evaluate the significance of a drug interaction.

This dissertation reviews the development and implementation of two systems biology meth- ods: ADVOCATE and hpARACNE. ADVOCATE was designed to deconvolve epithelium and stroma compartments fractions and virtual expression profiles from bulk gene expression profiles from human patients. We used laser capture microdissection and RNA sequencing to disentangle the transcriptional programs active in the malignant epithelium and stroma of pancreatic ductal adenocarcinoma (PDA), an aggressive malignancy with a prominent stromal component. We learned that distinct molecular subtypes are present in both the epithelium and the stroma of pancreatic cancer, and that the subtype identity of these two compartments are independent of one another. Critically, we discovered that specific com- binations of epithelial and stromal subtypes are strongly associated with patient survival across multiple external datasets, exhibiting both an effect-size and a level of reproducibility that was absent from previous efforts. These analyses were made possible by a new proba- bilistic algorithm (Adaptive DeconVolution Of CAncer Tissue Expression - ADVOCATE) that can extract compartment-specific gene expression profiles from bulk gene expression data. ADVOCATE accurately predicted the compartment fractions of bulk tumor samples and improved the performance of molecular classifiers by controlling for the diverse cellular compositions of independent datasets. This approach provides a much-needed framework to handle solid tumor tissue heterogeneity, allowing integrated analysis of both epithelial and stromal transcriptional programs from individual bulk samples. Reverse engineering approaches have been used to systematically dissect regulatory in- teractions based on gene expression profiles in different context and data types, thus im- proving our mechanistic understanding of molecular programs under perturbations. Pro- teomics data, on the other hand, provides direct evidence of cell functions. Particularly, signaling molecules are best candidates for drug targets. Previous efforts have shown that targeting signaling proteins could potentially lead to cancer remission. In this work, I introduce hybrid proteomics Algorithm for the Reconstruction of Accurate Cellular Network (hpARACNE), a re-design of gene expression based ARACNE algorithm. Us- ing Clinical Proteomics Tumor Analysis Consortium (CPTAC) breast cancer proteomics data, hpARACNE reconstructs a network that significantly outperforms ARACNE when compared with curated Kinase/Phosphatase-substrates interactions from public databases. Compared with Stable Isotope Labeling with Amino acid in Cell Culture (SILAC) ex- perimentally identified substrates for EGFR, hpARACNE predicts substrates with high accuracy. Integrative network analysis of breast cancer transcriptome and phosphopro- teome reveals potential drug targets for Triple Negative Breast Cancer (TNBC) treat- ment. hpARACNE has three innovations that adapt it to proteomics data and signaling process: 1) Refinement of the kinase/phosphatase peptides by integrating matched whole proteomic and whole phosphoproteomic profiles; 2) Establishment of association based on newly designed Mutual Information (MI) estimator for missing data; 3) Network pruning using directional Data Processing Inequality (dDPI) for signalling process.