Effectively developing robust predictive models for forecasting infectious disease dynamics over space and time relies on successful surveillance strategies to accurately assess host infection status. We are constantly refining these models to improve our understanding of transmission and persistence dynamics in host populations but are continuously challenged with difficulties in accurately diagnosing host infection status. These challenges are especially persistent for pathogens of the Mycobacterium tuberculosis Complex (MTBC), which cause tuberculosis (TB) disease in a wide array of mammalian hosts. These challenges are further exacerbated when working with MTBC pathogens in free-ranging wildlife hosts. Although TB disease in humans is a primary concern, TB in free-ranging wildlife hosts poses a large threat to human and animal health. One recently described and novel MTBC pathogen is Mycobacterium mungi, which infects the highly social, group-living banded mongoose (Mungos mungo). M. mungi poses a large threat to human and animal health as banded mongoose hosts thrive in urbanized areas and live in close proximity to humans, but despite this threat, accurately diagnosing M. mungi infection status remains a primary challenge. Here, I develop a host response-based assay for differentiating banded mongoose with clinical M. mungi disease from individuals that are putatively healthy using transcriptional biomarkers in whole blood. To our knowledge, this is the first evaluation of host response-based transcriptional signatures to detect TB infection in unstimulated whole blood collected from a free-ranging wildlife species. I found that the expression of two genes, GBP5 and DUSP3, are significantly upregulated (GBP5, p < .05; DUSP3, p < .005) in banded mongoose with clinical M. mungi disease when compared to that of putatively healthy individuals. These results are consistent with studies of active M. tuberculosis disease in humans and support the use of host response-based assays using blood transcriptional biomarkers for diagnosing TB in free-ranging wildlife hosts. These findings are important for improving surveillance strategies for diagnosing M. mungi infection status in banded mongoose and will be essential in refining predictive models for forecasting transmission and persistence dynamics over space and time. / Creating models to predict how diseases circulate and persist within a population is dependent on our ability to accurately diagnose if a host is infected. Diagnosing infection is difficult for some diseases, including tuberculosis (TB) pathogens, which infect humans and many other mammalian species. While vast improvements have been made in diagnosing TB infection in humans, diagnosing TB in free-ranging wildlife species is a constant challenge. These challenges are further exacerbated across the different pathogen species of TB. Although TB disease in humans is a primary concern, TB in free-ranging wildlife hosts poses a large threat to human and animal health. One recently discovered TB pathogen is Mycobacterium mungi, which infects free-ranging banded mongoose (Mungos mungo). This pathogen poses a large threat to human and animals health since banded mongoose thrive in urbanized areas and live in close proximity to humans. Despite this threat, accurately diagnosing M. mungi infection in banded mongoose remains a challenge. Here, I develop a diagnostic molecular tool that uses banded mongoose blood to measure the expression of specific genes and differentiate diseased individuals from seemingly healthy individuals. To our knowledge, this is the first study that has used this specific approach for diagnosing TB in a free-ranging wildlife species. I found that the expression of two genes are significantly increased in banded mongoose with clinical M. mungi disease when compared to that of seemingly healthy individuals. These results are consistent with studies human TB disease in humans and support the use of this approach for diagnosing TB in free-ranging wildlife hosts. These findings are important for improving diagnostics for M. mungi infection in banded mongoose and will be essential in refining models for predicting how this disease circulates and persists over space and time.

Quorum sensing in Gram-negative bacteria is best understood in the bioluminescent marine microorganism, <i>Vibrio fischeri</i>. In <i>V. fischeri</i>, the luminescence or <i>lux</i> genes are regulated in a cell density-dependent manner by the activator LuxR in the presence of an acylated homoserine lactone autoinducer molecule (3-oxo-hexanoyl homoserine lactone). LuxR, which binds to the <i>lux</i> operon promoter at position -42.5, is thought to function as an ambidextrous activator making multiple contacts with RNA polymerase (RNAP). The specific role of the <font face = "symbol">a</font>CTD of RNAP in LuxR-dependent transcriptional activation of the <i>lux</i> operon promoter has been investigated. The effect of seventy alanine substitution variants of the <font face = "symbol">a</font> subunit was determined <i>in vivo</i> by measuring the rate of transcription of the <i>lux</i> operon via luciferase assays in recombinant <i>Escherichia coli</i>. The mutant RNAPs from strains exhibiting at least two fold increased or decreased activity in comparison to the wild-type were further examined by <i>in vitro</i> assays. Since full-length LuxR has not been purified to date, an autoinducer-independent N-terminal truncated form of LuxR, LuxR<font face = "symbol">D</font>N, was used for <i>in vitro</i> studies. Single-round transcription assays were performed using reconstituted mutant RNAPs in the presence of LuxR<font face = "symbol">D</font>N, and fourteen residues in the <font face = "symbol">a</font>CTD were identified as having negative effects on the rate of transcription from the <i>lux</i> operon promoter. Five of these fourteen residues were also involved in the mechanism of both LuxR and LuxR<font face = "symbol">D</font>N-dependent activation <i>in vivo</i> and were chosen for further analysis by DNA mobility shift assays. Results from these assays indicate that while the wild-type <font face = "symbol">a</font>CTD is capable of interacting with the <i>lux</i> DNA fragment tested, all five of the variant forms of the <font face = "symbol">a</font>CTD tested appear to be deficient in their ability to recognize and bind the DNA. These findings suggest that <font face = "symbol">a</font>CTD-DNA interactions may play a role in LuxR-dependent transcriptional activation of the <i>lux</i> operon during quorum sensing. / Master of Science

Computational biology currently faces challenges in a big data world with thousands of data samples across multiple disease types including cancer. The challenging problem is how to extract biologically meaningful information from large-scale genomic data. Next-generation Sequencing (NGS) can now produce high quality data at DNA and RNA levels. However, in cells there exist a lot of non-specific (background) signals that affect the detection accuracy of true (foreground) signals. In this dissertation work, under Bayesian framework, we aim to develop and apply approaches to learn the distribution of genomic signals in each type of NGS data for reliable identification of specific foreground signals. We propose a novel Bayesian approach (ChIP-BIT) to reliably detect transcription factor (TF) binding sites (TFBSs) within promoter or enhancer regions by jointly analyzing the sample and input ChIP-seq data for one specific TF. Specifically, a Gaussian mixture model is used to capture both binding and background signals in the sample data; and background signals are modeled by a local Gaussian distribution that is accurately estimated from the input data. An Expectation-Maximization algorithm is used to learn the model parameters according to the distributions on binding signal intensity and binding locations. Extensive simulation studies and experimental validation both demonstrate that ChIP-BIT has a significantly improved performance on TFBS detection over conventional methods, particularly on weak binding signal detection. To infer cis-regulatory modules (CRMs) of multiple TFs, we propose to develop a Bayesian integration approach, namely BICORN, to integrate ChIP-seq and RNA-seq data of the same tissue. Each TFBS identified from ChIP-seq data can be either a functional binding event mediating target gene transcription or a non-functional binding. The functional bindings of a set of TFs usually work together as a CRM to regulate the transcription processes of a group of genes. We develop a Gibbs sampling approach to learn the distribution of CRMs (a joint distribution of multiple TFs) based on their functional bindings and target gene expression. The robustness of BICORN has been validated on simulated regulatory network and gene expression data with respect to different noise settings. BICORN is further applied to breast cancer MCF-7 ChIP-seq and RNA-seq data to identify CRMs functional in promoter or enhancer regions. In tumor cells, the normal regulatory mechanism may be interrupted by genome mutations, especially those somatic mutations that uniquely occur in tumor cells. Focused on a specific type of genome mutation, structural variation (SV), we develop a novel pattern-based probabilistic approach, namely PSSV, to identify somatic SVs from whole genome sequencing (WGS) data. PSSV features a mixture model with hidden states representing different mutation patterns; PSSV can thus differentiate heterozygous and homozygous SVs in each sample, enabling the identification of those somatic SVs with a heterozygous status in the normal sample and a homozygous status in the tumor sample. Simulation studies demonstrate that PSSV outperforms existing tools. PSSV has been successfully applied to breast cancer patient WGS data for identifying somatic SVs of key factors associated with breast cancer development. In this dissertation research, we demonstrate the advantage of the proposed distributional learning-based approaches over conventional methods for NGS data analysis. Distributional learning is a very powerful approach to gain biological insights from high quality NGS data. Successful applications of the proposed Bayesian methods to breast cancer NGS data shed light on underlying molecular mechanisms of breast cancer, enabling biologists or clinicians to identify major cancer drivers and develop new therapeutics for cancer treatment. / Ph. D.

This dissertation aims to develop a latent variable modeling framework with which to analyze gene expression profiling data for computational dissection of molecular signatures and transcriptional modules. The first part of the dissertation is focused on extracting pure gene expression signals from tissue or cell mixtures. The main goal of gene expression profiling is to identify the pure signatures of different cell types (such as cancer cells, stromal cells and inflammatory cells) and estimate the concentration of each cell type. In order to accomplish this, a new blind source separation method is developed, namely, nonnegative partially independent component analysis (nPICA), for tissue heterogeneity correction (THC). The THC problem is formulated as a constrained optimization problem and solved with a learning algorithm based on geometrical and statistical principles. The second part of the dissertation sought to identify gene modules from gene expression data to uncover important biological processes in different types of cells. A new gene clustering approach, nonnegative independent component analysis (nICA), is developed for gene module identification. The nICA approach is completed with an information-theoretic procedure for input sample selection and a novel stability analysis approach for proper dimension estimation. Experimental results showed that the gene modules identified by the nICA approach appear to be significantly enriched in functional annotations in terms of gene ontology (GO) categories. The third part of the dissertation moves from gene module level down to DNA sequence level to identify gene regulatory programs by integrating gene expression data and protein-DNA binding data. A sparse hidden component model is first developed for this problem, taking into account a well-known biological principle, i.e., a gene is most likely regulated by a few regulators. This is followed by the development of a novel computational approach, motif-guided sparse decomposition (mSD), in order to integrate the binding information and gene expression data. These computational approaches are primarily developed for analyzing high-throughput gene expression profiling data. Nevertheless, the proposed methods should be able to be extended to analyze other types of high-throughput data for biomedical research. / Ph. D.

The threat from infectious diseases dates as far back as prehistoric times. Pathogens continue to pose serious challenges to human health. The emergence and spread of diseases such as HIV/AIDS, Severe Acute Respiratory Syndrome (SARS), avian influenza, and the threats of bioterrorism have made infectious diseases major public health concerns. Despite many successes in the discovery of anti-infective medications, the treatment of infectious diseases faces serious challenges, which include (i) the emergence and reemergence of infectious pathogens, (ii) the ability of pathogens to adapt and develop resistance to drugs, and (ii) a shortage in the development and discovery of new anti-infective drugs. Host-Oriented Broad-Spectrum (HOBS) treatments have the promising potential to alleviate these problems. The HOBS treatment paradigm focuses on finding drug targets in human host that are simultaneously effective against a wide variety of infectious agents and toxins. In this dissertation, we present a computational approach to predict HOBS treatments by integrative analysis of three types of data, namely, (a) gene expression data representing host responses upon infection by a pathogen, (b) annotations of genes to pre-defined biological pathways and processes, and (iii) genes that are targets of known drugs. Our methods combine gene set-level enrichment with biclustering. We applied our approach to a compendium of gene expression data sets derived from host cells exposed to bacterial or to fungal pathogens, to functional annotation data from multiple databases, and to drug targets from DrugBank. We present putative host drug targets and drugs with extensive support in the literature for their potential to treat multiple bacterial and fungal infections. These results showcase the potential of our computational approach to predict HOBS drug targets that may be effective against two or more pathogens. Our study takes a clean-slate approach that promises to yield unsuspected or unknown associations between pathogens and biological processes, and thus discern candidate gene/proteins to be further probed as HOBS targets. Furthermore, by focusing on host responses to pathogens as captured by transcriptional data, our proposed approach stimulates host-oriented drug target identification, which has potential to alleviate the problem of drug resistance. / Ph. D.

Plants have molecular mechanisms for nutrient-related stress responses; however, their exact regulation remains unclear. For example, the integral myo-inositol (inositol) signal transduction pathway allows Arabidopsis thaliana to sense and respond to changes in environmental stimuli, such as water, light availability, and nutrient stress. The inositol signaling pathway relies on dynamic changes in second messenger levels of inositol(1,4,5)P3 (InsP3) and is regulated by myo-inositol polyphosphate 5-phosphatases (5PTases). The 5PTses keep balance between InsP3 signal transduction and termination. Previous work has identified the Sucrose non-fermenting (Snf) 1-related kinase (SnRK1.1) as a binding partner to 5PTase13, a potential InsP3 regulator, and a novel protein called P80, a predicted component of the Cullin4 (CUL4) E3 Ubiquitin ligase complex. In plants, SnRK1.1 is a central integrator of metabolism, stress responses, and developmental signals. Moreover, SnRK1.1 is conserved with the eukaryotic AMP-activated protein (AMPK) and Snf1 kinases—enzymes fundamental to transcriptional regulation and metabolic balance. Studying SnRK1.1 regulation may reveal mechanisms for agricultural sustainability and may offer valuable links to understanding metabolic diseases and lifespan in humans. Therefore, the research presented here centered on characterizing the regulation of SnRK1 gene expression and steady-state protein levels in plants. I show developmental and nutrient-related regulation of spatial expression patterns of SnRK1 genes and SnRK1.1 protein. Further, I present a model for regulation of SnRK1.1 protein stability in vivo based on SnRK1.1 steady-state protein levels in p80 and cul4 co-suppressed (cs) mutants. My results indicate SnRK1.1 regulation is dynamic, and dependent on the timing of particular cues from development and the environment. / Master of Science

Gene expression regulation encompasses a wide range of mechanisms that govern cellular processes. Among these, post-transcriptional regulation, including translation control, plays a pivotal role in ensuring precise protein synthesis, timing, and quantity. Perturbations of mechanisms such as RNA modifications, and interactions between RNA-binding proteins (RBPs) and specific RNA motifs, can lead to dysregulation of essential cellular processes. These alterations contribute to the development of various disorders, including cancer, neurodegenerative diseases, and metabolic disorders. Many publicly available datasets and studies offer opportunities to investigate the link between alterations in these mechanisms and disease manifestations. However, the limited availability of datasets for certain conditions or notable inconsistencies among reported associations prevent complete understanding of the underlying processes. Therefore, extending the investigations to encompass a diverse range of genes and/or diseases will enhance our comprehension of these intricate regulatory and disease mechanisms, aiding in the identification of potential therapeutic targets and innovative interventions to mitigate pathological conditions. In particular, we focused on three separate aspects involved in gene expression regulation: RNA modifications, RBPs interactions with RNA secondary structures, and the Kozak consensus sequence as a translational modulator. Each part uncovers essential mechanisms that govern post-transcriptional control of gene expression, shedding light on their roles in cellular processes and disease contexts. At first, we performed a comprehensive exploration of 15 RBPs involved in the regulation of the N6-methyladenosine (m6A) methylation. Leveraging data from The Cancer Genome Atlas (TCGA), we conducted a pan-cancer analysis across 31 tumor types to uncover the distribution of alterations of these factors, and we developed a user-friendly web application to enable users to conduct similar analyses. Additionally, we performed a parallel analysis focused on neuroblastoma, using data from publicly available and in-house datasets. These investigations unveil the potential impact of a subset of m6A factors on cancer development and progression. While in the first case, VIRMA and YTHDF reader proteins, emerged as the most frequently altered genes with significant pan-cancer prognostic implications, in the context of neuroblastoma, the writer METTL14 and the reader ALKBH5, showed the most prominent roles. Subsequently, our focus shifted to a distinct subset of RBPs capable of interacting with RNA secondary structures, particularly with RNA G-quadruplexes (RG4s). We established a comprehensive database cataloging RBPs with potential RG4-binding capabilities. This resource represents a valuable tool for researchers aiming to explore the intricate interplays between RBPs and RG4s, and their putative implications in diverse biological processes and diseases. Finally, attention was directed to the Kozak sequence, a pivotal determinant of the regulation of translation initiation. Exploiting the power of base editors, we developed a method to optimize translation initiation by modifying the Kozak sequence. This strategy offers promise in addressing haploinsufficiency-related disorders, where enhancing the functional protein is essential. Overall, these findings present opportunities for the identification of potential therapeutic targets and precision medicine strategies to alleviate a spectrum of pathological conditions, thus fostering advancements in the field of molecular biology and disease management.

Recent breakthroughs in high-throughput biotechnology have promoted the integration of multi-platform data to investigate signal transduction pathways within a cell. In order to model complicated dynamics and heterogeneity of biological pathways, sophisticated computational models are needed to address unique properties of both the biological hypothesis and the data. In this dissertation work, we have proposed and developed methods using Markov Chain Monte Carlo (MCMC) techniques to solve complex modeling problems in human cancer research by integrating multi-platform data. We focus on two research topics: 1) identification of transcriptional regulatory networks and 2) uncovering of aberrant intracellular signal transduction pathways. We propose a robust method, called GibbsOS, to identify condition specific gene regulatory patterns between transcription factors and their target genes. A Gibbs sampler is employed to sample target genes from the marginal function of outlier sum of regression t statistic. Numerical simulation has demonstrated significant performance improvement of GibbsOS over existing methods against noise and false positive connections in binding data. We have applied GibbsOS to breast cancer cell line datasets and identified condition specific regulatory rewiring in human breast cancer. We also propose a novel method, namely Gibbs sampler to Infer Signal Transduction (GIST), to detect aberrant pathways that are highly associated with biological phenotypes or clinical information. By converting predefined potential functions into a Gibbs distribution, GIST estimates edge directions by learning the distribution of linear signaling pathway structures. Through the sampling process, the algorithm is able to infer signal transduction directions which are jointly determined by both gene expression and network topology. We demonstrate the advantage of the proposed algorithms on simulation data with respect to different settings of noise level in gene expression and false-positive connections in protein-protein interaction (PPI) network. Another major contribution of the dissertation work is that we have improved traditional perspective towards understanding aberrant signal transductions by further investigating structural linkage of signaling pathways. We develop a method called Structural Organization to Uncover pathway Landscape (SOUL), which emphasizes on modularized pathways structures from reconstructed pathway landscape. GIST and SOUL provide a very unique angle to computationally model alternative pathways and pathway crosstalk. The proposed new methods can bring insight to drug discovery research by targeting nodal proteins that oversee multiple signaling pathways, rather than treating individual pathways separately. A complete pathway identification protocol, namely Infer Modularization of PAthway CrossTalk (IMPACT), is developed to bridge downstream regulatory networks with upstream signaling cascades. We have applied IMPACT to breast cancer treated patient datasets to investigate how estrogen receptor (ER) signaling pathways are related to drug resistance. The identified pathway proteins from patient datasets are well supported by breast cancer cell line models. We hypothesize from computational results that HSP90AA1 protein is an important nodal protein that oversees multiple signaling pathways to drive drug resistance. Cell viability analysis has supported our hypothesis by showing a significant decrease in viability of endocrine resistant cells compared with non-resistant cells when 17-AAG (a drug that inhibits HSP90AA1) is applied. We believe that this dissertation work not only offers novel computational tools towards understanding complicated biological problems, but more importantly, it provides a valuable paradigm where systems biology connects data with hypotheses using computational modeling. Initial success of using microarray datasets to study endocrine resistance in breast cancer has shed light on translating results from high throughput datasets to biological discoveries in complicated human disease studies. As the next generation biotechnology becomes more cost-effective, the power of the proposed methods to untangle complicated aberrant signaling rewiring and pathway crosstalk will be finally unleashed. / Ph. D.

The mechanism of gene regulation used by Gram-negative bacteria during quorum sensing is well understood in the bioluminescent marine bacterium Vibrio fischeri. The cell-density dependent activation of the luminescence (lux) genes of V. fischeri relies on the formation of a complex between the autoinducer molecule, N-(3-oxohexanoyl) homoserine lactone, and the autoinducer-dependent transcriptional activator LuxR. LuxR, a 250 amino acid polypeptide, binds to a site known as the lux box centered at position -42.5 relative to the luxI transcriptional start site. During transcriptional activation of the lux operon, LuxR is thought to function as an ambidextrous activator capable of making multiple contacts with RNA polymerase (RNAP). The specific role of region 4 of the Escherichia coli sigma 70 subunit of RNAP in LuxR-dependent transcriptional activation of the luxI promoter has been investigated. Rich in basic amino acids, this conserved portion of sigma 70 is likely to be surface-exposed and available to interact with transcription factors bound near the -35 element. The effect of 16 single and 2 triple alanine substitution variants of sigma 70 between amino acid residues 590 and 613, was determined in vivo by measuring the rate of transcription from a luxI-lacZ translational fusion via b-galactosidase assays in recombinant E. coli. In vitro work was performed with LuxRDN, the autoinducer-independent C-terminal domain (amino acids 157 to 250) of LuxR because purified, full length LuxR is unavailable. Single-round transcription assays were performed in the presence of LuxRDN and 19 variant RNAPs, one of which contained a C-terminally truncated sigma 70 subunit devoid of region 4. Results indicate that region 4 is essential for LuxRDN-dependent luxI transcription with two specific amino acid residues, E591 and K597, having negative effects on the rate of LuxRDN-dependent luxI transcription in vivo and in vitro. None of the residues tested were identified as having any effect on LuxR-dependent luxI transcription in vivo. These findings suggest that region 4.2 is most likely to be in close proximity to LuxR when bound to the luxI promoter. However, unlike the situation found for other ambidextrous activators, no single residue within region 4.2 of sigma 70 may be critical by itself for LuxR-dependent during transcriptional activation. / Master of Science

<I>Vibrio fischeri</I>, a symbiotic bioluminescent bacterium, serves as one of the best understood model systems for a mechanism of cell-density dependent bacterial gene regulation known as quorum sensing. During quorum sensing in <I>V. fischeri</I>, an acylated homoserine chemical signal (autoinducer) is synthesized by the bacteria and used to sense their own species in a given environment. As the autoinducer levels rise, complexes form between the autoinducer and the N-terminal domain of a regulatory protein, LuxR. In response to autoinducer binding, LuxR is believed to undergo a conformational change that allows the C-terminal domain to activate transcription of the luminescence or <I>lux</I> operon. To further understand the mechanism of LuxR-dependent transcriptional activation of the <I>lux</I> operon, PCR-based site-directed mutagenesis procedures have been used to generate alanine-substitution mutants in the C-terminal forty-one amino acid residues of LuxR, a region that has been hypothesized to play a critical role in the activation process. An <I>in vivo</I> luminescence assay was first used to test the effects of the mutations on LuxR-dependent activation of the <I>lux</I> operon in recombinant <I>Escherichia coli</I>. Luciferase levels present in cell extracts obtained from these strains were also quantified and found to correlate with the luminescence results. Eight strains encoding altered forms of LuxR exhibited a "dark" phenotype with luminescence output less than 50% and luciferase levels less than 50% of the wildtype control strain. Western immunoblotting analysis with cell extracts from the luminescence and luciferase assays verified that the altered forms of LuxR were expressed at levels approximately equal to wildtype. Therefor, Low luminescence and luciferase levels could be the result of a mutation that either affects the ability of LuxR to recognize and bind its DNA target (the <I>lux</I> box) or to establish associations with RNA polymerase (RNAP) at the <I>lux</I> operon promoter necessary for transcriptional initiation. A third <I>in vivo </I>assay was used to test the ability of the altered forms of LuxR to bind to the <I>lux</I> box (DNA binding assay/repression). All of the LuxR variants exhibiting the "dark" phenotype in the luminescence and luciferase assay were also found to be unable to bind to the <I>lux</I> box in the<I> </I>DNA binding assay. Therefore, it can be concluded that the alanine substitutions made at these positions affect the ability of LuxR to bind to the <I>lux</I> box in the presence and absence of RNA polymerase. Another class of mutants exhibited wildtype phenotypes in the luminescence and luciferase assays but were unable to bind to the <I>lux</I> box in the DNA binding assay. The alanine substitutions made at these amino acid residues may be making contacts with RNAP that are important for maintaining the stability of the DNA binding region of LuxR. Alanine substitutions made at these positions have a defect in DNA binding at the promoter of the <I>lux</I> operon only in the absence of RNAP. None of the alanine substitutions made in the C-terminal forty-one amino acids of LuxR were found to affect activation of transcription of the <I>lux</I> operon without also affecting DNA binding. Taken together, these results support the conclusion that the C-terminal forty-one amino acids of LuxR are important for DNA recognition and binding of the <I>lux</I> box rather than positive control of the process of transcription initiation. / Master of Science