No description available.

Durante las últimas décadas el desarrollo de la bioinformática nos ha permitido lograr una mayor comprensión de los procesos biológicos que ocurren con nuestras células a nivel molecular. Al respecto, las mejoras e innovaciones en la tecnología continúan estimulando la mejora en la calidad de los datos biológicos que pueden ser obtenidos a nivel genómico. En tal sentido, grandes volúmenes de información pueden ser encontrados en formas de anotaciones o bases de datos computacionales. Estos conjuntos de datos, apropiadamente combinados, tienen el potencial de posibilitar descubrimientos novedosos que lleven a avances en campos tan relevantes para el desarrollo nacional como son la biotecnología o la medicina post-genómica. En particular, esta tesis se centra en la investigación de técnicas de aprendizaje automático y computación evolutiva para la inferencia de redes regulatorias de genes a partir de datos de expresión de genes, a nivel de genomas completos. Una red regulatoria de genes es una colección de segmentos de ADN (ácido desoxirribonucleico) en una célula que interactúan unos con otros (indirectamente a través del producto de su expresión) y con otras sustancias en la célula, gobernando así las tasas de transcripción de los genes de la red en ARNm (ácido ribonucleico mensajero). La principal contribución de esta tesis esta relacionada con el desarrollo de metodologías computacionales que asistan, a expertos en bioinformática, en la ingeniería inversa de las redes regulatorias de genes. En tal sentido, se desarrollaron algoritmos de computación evolutiva que permiten la identificación de grupos de genes co-expresados bajo ciertos subconjuntos de condiciones experimentales. Estos algoritmos se aplican sobre datos de expresión de genes, y optimizan características deseables desde el punto de vista biológico, posibilitando la obtención de relaciones de co-expresión relevantes. Tales algoritmos fueron cuidadosamente validados por medio de comparaciones con otras técnicas similares disponibles en la literatura, realizando estudios con datos reales y sintéticos a fin de mostrar la utilidad de la información extraída. Además, se desarrolló un algoritmo de inferencia que permite la extracción de potenciales relaciones causa-efecto entre genes, tanto simultáneas como también aquellas diferidas en el tiempo. Este algoritmo es una evolución de una técnica presentada con anterioridad, e incorpora características novedosas como la posibilidad de inferir reglas con múltiples retardos en el tiempo, a nivel genoma completo, e integrando múltiples conjuntos de datos. La técnica se validó mostrando su eficacia respecto de otros enfoques relevantes de la literatura. También se estudiaron los resultados obtenidos a partir de conjuntos de datos reales en términos de su relevancia biológica, exponiendo la viabilidad de la información inferida. Finalmente, estos algoritmos se integraron en una plataforma de software que facilita la utilización de estas técnicas permitiendo la inferencia, manipulación y visualización de redes regulatorias de genes. / In recent decades, the development of bioinformatics has allowed us to achieve a greater understanding of the biological processes that occur at the molecular level in our cells. In this regard, the improvements and innovations in technology continue to boost the improvement in the quality of the biological data that can be obtained at the genomic level. In this regard, large volumes of information can be found in forms of ontology's or computer databases. These datasets, appropriately combined, have the potential to enable novel discoveries that lead to progress in relevant fields to national development such as biotechnology and post-genomic medicine. In particular, this thesis focuses on the research of machine learning techniques and evolutionary computation for the inference of gene regulatory networks from gene expression data at genome-wide levels. A gene regulatory network is a collection of segments of DNA (deoxyribonucleic acid) in a cell which interact with each other (indirectly through their products of expression) and with other substances in the cell, thereby governing the rates of network genes transcription into mRNA (messenger ribonucleic acid). The main contribution of this thesis is related to the development of computational methodologies to attend experts in bioinformatics in the reverse engineering of gene regulatory networks. In this sense, evolutionary algorithms that allow the identification of groups of coexpressed genes under certain subsets of experimental conditions were developed. These algorithms are applied to gene expression data, and optimize desirable characteristics from the biological point of view, allowing the inference of relevant co-expression relationships. Such algorithms were carefully validated by the comparison with other similar techniques available in the literature, conducting studies with real and synthetic data in order to show the usefulness of the information extracted. Furthermore, an inference algorithm that allows the extraction of potential cause-effect relationships between genes, both simultaneous and time-delayed, were developed. This algorithm is an evolution of a previous approach, and incorporates new features such as the ability to infer rules with multiple time delays, at genome-wide level, and integrating multiple datasets. The technique was validated by showing its effectiveness over other relevant approaches in the literature. The results obtained from real datasets were also studied in terms of their biological relevance by exposing the viability of the inferred information. Finally, these algorithms were integrated into a software platform that facilitates the use of these techniques allowing the inference, manipulation and visualization of gene regulatory networks.

One goal of systems biology is to predict and modify the behavior of biological networks by accurately monitoring and modeling their responses to certain types of perturbations. The construction of mathematical models based on observation of these responses, referred to as reverse engineering, is an important step in elucidating the structure and dynamics of such networks. Continuous models, described by systems of differential equations, have been used to reverse engineer biochemical networks. Of increasing interest is the use of discrete models, which may provide a conceptual description of the network. In this dissertation we introduce a discrete modeling approach, rooted in computational algebra, to reverse-engineer networks from experimental time series data. The algebraic method uses algorithmic tools, including Groebner-basis techniques, to build the set of all discrete models that fit time series data and to select minimal models from this set. The models used in this work are discrete-time finite dynamical systems, which, when defined over a finite field, are described by systems of polynomial functions. We present novel reverse-engineering algorithms for discrete models, where each algorithm is suitable for different amounts and types of data. We demonstrate the effectiveness of the algorithms on simulated networks and conclude with a description of an ongoing project to reverse-engineer a real gene regulatory network in yeast. / Ph. D.

Mathematical models of regulatory networks become more difficult to construct and understand as they grow in size and complexity. Large regulatory network models can be built up from smaller models, representing subsets of reactions within the larger network. This dissertation focuses on novel model construction techniques that extend the ability of biological modelers to construct larger models by supplying them with tools for decomposing models and using the resulting components to construct larger models. Over the last 20 years, molecular biologists have amassed a great deal of information about the genes and proteins that carry out fundamental biological processes within living cells --- processes such as growth and reproduction, movement, signal reception and response, and programmed cell death. The full complexity of these macromolecular regulatory networks is too great to tackle mathematically at the present time. Nonetheless, modelers have had success building dynamical models of restricted parts of the network. Systems biologists need tools now to support composing "submodels" into more comprehensive models of integrated regulatory networks. We have identified and developed four novel processes (fusion, composition, flattening, and aggregation) whose purpose is to support the construction of larger models. Model Fusion combines two or more models in an irreversible manner. In fusion, the identities of the original (sub)models are lost. Beyond some size, fused models will become too complex to grasp and manage as single entities. In this case, it may be more useful to represent large models as compositions of distinct components. In Model Composition one thinks of models not as monolithic entities but rather as collections of smaller components (submodels) joined together. A composed model is built from two or more submodels by describing their redundancies and interactions. While it is appealing in the short term to build larger models from pre-existing models, each developed independently for their own purposes, we believe that ultimately it will become necessary to build large models from components that have been designed for the purpose of combining them. We define Model Aggregation as a restricted form of composition that represents a collection of model elements as a single entity (a "module"). A module contains a definition of pre-determined input and output ports. The process of aggregation (connecting modules via their interface ports) allows modelers to create larger models in a controlled manner. Model Flattening converts a composed or aggregated model with some hierarchy or connections to one without such connections. The relationships used to describe the interactions among the submodels are lost, as the composed or aggregated model is converted into a single large (flat) model. Flattening allows us to use existing simulation tools, which have no support for composition or aggregation. / Ph. D.

Mathematical modeling has become increasingly popular as a tool to study regulatory interactions within gene-protein networks. From the modeler's perspective, two challenges arise in the process of building a mathematical model. First, the same regulatory network can be translated into different types of models at different levels of detail, and the modeler must choose an appropriate level to describe the network. Second, realistic regulatory networks are complicated due to the large number of biochemical species and interactions that govern any physiological process. Constructing and validating a realistic mathematical model of such a network can be a difficult and lengthy task. To confront the first challenge, we develop a new modeling approach that classifies components in the networks into three classes of variables, which are described by different rate laws. These three classes serve as "building blocks" that can be connected to build a complex regulatory network. We show that our approach combines the best features of different types of models, and we demonstrate its utility by applying it to the budding yeast cell cycle. To confront the second challenge, modelers have developed rule-based modeling as a framework to build complex mathematical models. In this approach, the modeler describes a set of rules that instructs the computer to automatically generate all possible chemical reactions in the network. Building a mathematical model using rule-based modeling is not only less time-consuming and error-prone, but also allows modelers to account comprehensively for many different mechanistic details of a molecular regulatory system. We demonstrate the potential of rule-based modeling by applying it to the generation of circadian rhythms in cyanobacteria. / Ph. D.

A gene regulatory network (GRN) records the interactions among transcription factors and their target genes. GRNs are useful to study how transcription factors (TFs) control gene expression as cells transition between states during differentiation and development. Scientists usually construct GRNs by careful examination and study of the literature. This process is slow and painstaking and does not scale to large networks. In this thesis, we study the problem of inferring GRNs automatically from gene expression data. Recent data-driven approaches to infer GRNs increasingly rely on single-cell level RNA-sequencing (scRNA-seq) data. Most of these methods rely on unsupervised or association based strategies, which cannot leverage known regulatory interactions by design. To facilitate supervised learning, we propose a novel graph convolutional neural network (GCN) based autoencoder to infer new regulatory edges from a known GRN and scRNA-seq data. As the name suggests, a GCN-based autoencoder consists of an encoder that learns a low-dimensional embedding of the nodes (genes) in the input graph (the GRN) through a series of graph convolution operations and a decoder that aims to reconstruct the original graph as accurately as possible. We investigate several GCN-based architectures to determine the ideal encoder-decoder combination for GRN reconstruction. We systematically study the performance of these and other supervised learning methods on different mouse and human scRNA-seq datasets for two types of evaluation. We demonstrate that our GCN-based approach substantially outperforms traditional machine learning approaches. / Master of Science / In multi-cellular living organisms, stem cells differentiate into multiple cell types. Proteins called transcription factors (TFs) control the activity of genes to effect these transitions. It is possible to represent these interactions abstractly using a gene regulatory network (GRN). In a GRN, each node is a TF or a gene and each edge connects a TF to a gene or TF that it controls. New high-throughput technologies that can measure gene expression (activity) in individual cells provide rich data that can be used to construct GRNs. In this thesis, we take advantage of recent advances in the field of machine learning to develop a new computational method for computationally constructing GRNs. The distinguishing property of our technique is that it is supervised, i.e., it uses experimentally-known interactions to infer new regulatory connections. We investigate several variations of this approach to reconstruct a GRN as close to the original network as possible. We analyze and provide a rationale for the decisions made in designing, evaluating, and choosing the characteristics of our predictor. We show that our predictor has a reconstruction accuracy that is superior to other supervised-learning approaches.

So called discrete models have been successfully used in engineering and computational systems biology. This thesis discusses algebraic methods for modeling and analysis of gene regulatory networks within the discrete modeling context. The first chapter gives a background for discrete models and put in context some of the main research problems that have been pursued in this field for the last fifty years. It also outlines the content of each subsequent chapter. The second chapter focuses on the problem of inferring dynamics from the structure (topology) of the network. It also discusses the characterization of the attractor structure of a network when a particular class of functions control the nodes of the network. Chapters~3 and 4 focus on the study of multi-state nested canalyzing functions as biologically inspired functions and the characterization of their dynamics. Chapter 5 focuses on stochastic methods, specifically on the development of a stochastic modeling framework for discrete models. Stochastic discrete modeling is an alternative approach from the well-known mathematical formalizations such as stochastic differential equations and Gillespie algorithm simulations. Within the discrete setting, a framework that incorporates propensity probabilities for activation and degradation is presented. This approach allows a finer analysis of discrete models and provides a natural setup for cell population simulations. Finally, Chapter 6 discusses future research directions inspired by the work presented here. / Ph. D.

The life cycle of Caulobacter crescentus is of interest because of the asymmetric nature of cell division that gives rise to progeny that have distinct morphology and function. One daughter called the stalked cell is sessile and capable of DNA replication, while the second daughter called the swarmer cell is motile but quiescent. Advances in microscopy combined with molecular biology techniques have revealed that macromolecules are localized in a non-homogeneous fashion in the cell cytoplasm, and that dynamic localization of proteins is critical for cell cycle progression and asymmetry. However, the molecular-level mechanisms that govern protein localization, and enable the cell to exploit subcellular localization towards orchestrating an asymmetric life cycle remain obscure. There are also instances of researchers using intuitive reasoning to develop very different verbal explanations of the same biological process. To provide a complementary view of the molecular mechanism controlling the asymmetric division cycle of Caulobacter, we have developed a mathematical model of the cell cycle regulatory network. Our reaction-diffusion models provide additional insight into specific mechanism regulating different aspects of the cell cycle. We describe a molecular mechanism by which the bifunctional histidine kinase PleC exhibits bistable transitions between phosphatase and kinase forms. We demonstrate that the kinase form of PleC is crucial for both swarmer-to-stalked cell morphogenesis, and for replicative asymmetry in the predivisional cell. We propose that localization of the scaffolding protein PopZ can be explained by a Turing-type mechanism. Finally, we discuss a preliminary model of ParA- dependent chromosome segregation. Our model simulations are in agreement with experimentally observed protein distributions in wild-type and mutant cells. In addition to predicting novel mutants that can be tested in the laboratory, we use our models to reconcile competing hypotheses and provide a unified view of the regulatory mechanisms that direct the Caulobacter cell cycle. / Ph. D.

I present novel solutions to two different classes of computational problems that arise in the study of complex cellular processes. The first problem arises in the context of planning large-scale genetic cross experiments that can be used to validate predictions of multigenic perturbations made by mathematical models. (i) I present CrossPlan, a novel methodology for systematically planning genetic crosses to make a set of target mutants from a set of source mutants. CrossPlan is based on a generic experimental workflow used in performing genetic crosses in budding yeast. CrossPlan uses an integer-linear-program (ILP) to maximize the number of target mutants that we can make under certain experimental constraints. I apply it to a comprehensive mathematical model of the protein regulatory network controlling cell division in budding yeast. (ii) I formulate several natural problems related to efficient synthesis of a target mutant from source mutants. These formulations capture experimentally-useful notions of verifiability (e.g., the need to confirm that a mutant contains mutations in the desired genes) and permissibility (e.g., the requirement that no intermediate mutants in the synthesis be inviable). I present several polynomial time or fixed-parameter tractable algorithms for optimal synthesis of a target mutant for special cases of the problem that arise in practice. The second problem I address is inferring gene regulatory networks (GRNs) from single cell transcriptomic (scRNA-seq) data. These GRNs can serve as starting points to build mathematical models. (iii) I present BEELINE, a comprehensive evaluation of state-of-the-art algorithms for inferring gene regulatory networks (GRNs) from single-cell gene expression data. The evaluations from BEELINE suggest that the area under the precision-recall curve and early precision of these algorithms are moderate. Techniques that do not require pseudotime-ordered cells are generally more accurate. Based on these results, I present recommendations to end users of GRN inference methods. BEELINE will aid the development of gene regulatory network inference algorithms. (iv) Based on the insights gained from BEELINE, I propose a novel graph convolutional neural network (GCN) based supervised algorithm for GRN inference form single-cell gene expression data. This GCN-based model has a considerably better accuracy than existing supervised learning algorithms for GRN inference from scRNA-seq data and can infer cell-type specific regulatory networks. / Doctor of Philosophy / A small number of key molecules can completely change the cell's state, for example, a stem cell differentiating into distinct types of blood cells or a healthy cell turning cancerous. How can we uncover the important cellular events that govern complex biological behavior? One approach to answering the question has been to elucidate the mechanisms by which genes and proteins control each other in a cell. These mechanisms are typically represented in the form of a gene or protein regulatory network. The resulting networks can be modeled as a system of mathematical equations, also known as a mathematical model. The advantage of such a model is that we can computationally simulate the time courses of various molecules. Moreover, we can use the model simulations to predict the effect of perturbations such as deleting one or more genes. A biologist can perform experiments to test these predictions. Subsequently, the model can be iteratively refined by reconciling any differences between the prediction and the experiment. In this thesis I present two novel solutions aimed at dramatically reducing the time and effort required for this build-simulate-test cycle. The first solution I propose is in prioritizing and planning large-scale gene perturbation experiments that can be used for validating existing models. I then focus on taking advantage of the recent advances in experimental techniques that enable us to measure gene activity at a single-cell resolution, known as scRNA-seq. This scRNA-seq data can be used to infer the interactions in gene regulatory networks. I perform a systematic evaluation of existing computational methods for building gene regulatory networks from scRNA-seq data. Based on the insights gained from this comprehensive evaluation, I propose novel algorithms that can take advantage of prior knowledge in building these regulatory networks. The results underscore the promise of my approach in identifying cell-type specific interactions. These context-specific interactions play a key role in building mathematical models to study complex cellular processes such as a developmental process that drives transitions from one cell type to another

No / Skin development, postnatal growth and regeneration are governed by complex and well-balanced programs of gene activation and silencing. The crosstalk between small non-coding microRNAs (miRNAs) and mRNAs is highly important for steadiness of signal transduction and transcriptional activities as well as for maintenance of homeostasis in many organs, including the skin. Recent data demonstrated that the expression of many genes, including cell type-specific master transcription regulators implicated in the control of skin development and homeostasis, is regulated by miRNAs. In addition, individual miRNAs could mediate the effects of these signaling pathways through being their downstream components. In turn, the expression of a major constituent of the miRNA processing machinery, Dicer, can be controlled by cell type-specific transcription factors, which form negative feedback loop mechanisms essential for the proper execution of cell differentiation- associated gene expression programs and cell-cell communications during normal skin development and regeneration. This review summarizes the available data on how miRNA/mRNA regulatory networks are involved in the control of skin development, epidermal homeostasis, hair cycle-associated tissue remodeling and pigmentation. Understanding of the fundamental mechanisms that govern skin development and regeneration will contribute to the development of new therapeutic approaches for many pathological skin conditions by using miRNA-based interventions.