Master of Science, Miami University, 2009, Botany

Gene duplication can lead to increased morphological complexity by giving rise to new genes and gene functions. The classic Ohno model predicts new gene duplicates will be functionally redundant. However, this model does not address functional divergence due to differential expression. To test this possibility, the developmental expression patterns of 18 pairs of unlinked recent duplicates in Arabidopsis thaliana were characterized. These duplicates have a Ks less than 5%, are less than 2 my old, and are specific to the A. thaliana lineage. The coding regions of these duplicates are generally complete, although they share only an average 616 bp of upstream regulatory sequence. We found that the majority of duplicate loci have non-redundant expression patterns, suggesting functional divergence occurs soon after duplication for unlinked duplicates. Furthermore, comparison of expression data from this study to online databases suggests these databases cannot reliably differentiate expression patterns among recent duplicate genes.

Committee: Richard Moore Dr. (Advisor); Qingshun Li Dr. (Committee Member); Martin Stevens Dr. (Committee Member) Subjects: Botany; Genetics; Molecular Biology

Doctor of Philosophy (PhD), Wright State University, 2022, Biomedical Sciences PhD

Gene expression in mammalian cells requires complex nuclear choreography, and there is increasing evidence that spatiotemporal organization of chromatin and nuclear compartments plays an important role in gene expression. In this dissertation, I examined the function of SON, a splicing factor with a known role in nuclear organization, in chromatin-mediated gene expression. SON association with a transcriptionally inactive U2OS 2-6-3 reporter gene array provided a useful model to study SON's chromatin dynamics. I demonstrated that SON associates with the inactive but not the activated array, and that SON's RNA binding domains are not necessary for that association. Second, I discovered a new role for SON in maintaining chromatin condensation. Whereas chromatin decondensation is typically correlated with transcription activation, I have demonstrated that reporter transcripts are not produced at decondensed SON-depleted U2OS 2-6-3 reporter gene loci, and that SON-depleted loci contain histone H3 that is trimethylated on lysine 9, a marker for transcriptionally silent chromatin. We found that SON depleted reporter loci are still transcriptionally activatable, and that inhibiting transcription elongation is not sufficient to condense the enlarged SON depleted reporter locus. These findings suggest that higher order chromatin structure and transcription activation are functionally distinct mechanisms of gene regulation that can be uncoupled. Finally, I investigated SON's role in genome-wide chromatin organization. SON-depleted cells are more susceptible to DNase digestion, implicating SON in the maintenance of chromatin stability globally. In conclusion, this study demonstrates a new function for the splicing factor SON in maintaining chromatin organization.

Committee: Paula Bubulya Ph.D. (Advisor); Quan Zhong Ph.D. (Committee Member); Labib Rouhana Ph.D. (Committee Member); Weiwen Long Ph.D. (Committee Member); Michael Leffak Ph.D. (Committee Member) Subjects: Biology; Biomedical Research; Cellular Biology; Molecular Biology

Master of Science in Biomedical Engineering (MSBME), Wright State University, 2019, Biomedical Engineering

Triple-negative breast cancer (TNBC) is characterized by the absence of expression of the estrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 (HER2). Therefore, TNBC is unresponsive to targeted hormonal therapies, which limits treatment options to nonselective chemotherapeutic agents. Basal-like breast cancers (BLBCs) represent a subset of about 70% of TNBCs, more frequently affecting younger patients, being more prevalent in African-American women and significantly more aggressive than tumors of other molecular subtypes, with high rates of proliferation and extremely poor clinical outcomes. Proper classification of BLBCs using current pathological tools has been a major challenge. Although TNBCs have many BLBC characteristics, the relationship between clinically defined TNBC and the gene expression profile of BLBC is not fully examined. The purpose of this study is to assemble publicly-available TNBC gene expression datasets generated by Affymetrix gene chips and define a set of genes, or gene signature, that can classify TNBC samples between BLBC and Non-BLBC subtypes. We compiled over 3,500 breast cancer gene expression profiles from several individual publicly available datasets and extracted Affymetrix gene expression data for 580 TNBC cases. Several popular data mining methods along with dimensionality reduction and feature selection techniques were applied to the resultant dataset to build predictive models to understand molecular characteristics and mechanisms associated with BLBCs and to classify them more accurately according to important features extracted through microarray data analysis of BLBC and Non-BLBC cases. Our result can lead to proper identification and diagnosis of BLBCs, which can potentially direct clinical implications by dictating the most effective therapy.

Committee: Jaime Ramirez-Vick Ph.D. (Advisor); Nasim Nosoudi Ph.D. (Committee Member); Amir Zadeh Ph.D. (Committee Member) Subjects: Biomedical Engineering; Genetics

Doctor of Philosophy, The Ohio State University, 2004, Veterinary Biosciences

Human T-lymphotropic virus type-1 (HTLV-1), a deltaretrovirus cause adult T-cell leukemia/lymphoma and other lymphocyte-mediated disorders. HTLV-1 provirus encodes various regulatory and accessory genes, including p30(II). Our laboratory has identified the functional properties of pX ORF-II encoded p30(II), but the role of this viral protein in virus replication or pathogenesis remains unclear. We have reported that HTLV-1 p30(II), a nuclear-localizing protein, interacts with CBP/p300, disrupts CREB-Tax-CBP/p300 complex formation at HTLV-1 Tax-Responsive Elements repeats (TRE) and differentially modulates CREB and TRE-mediated transcription. We have also recently demonstrated that p30(II) is critical in maintaining viral loads in vivo. Herein, we further characterized the role of p30(II) in regulation of cellular gene expression, using microarrays and identified alterations of interrelated pathways of cell-proliferation, T-cell signaling, apoptosis and cell-cycle in p30(II)-expressing T-lymphocytes. We are the first to test the overall effect of an HTLV-1 accessory protein, on cellular gene expression and demonstrate that p30(II) activates transcription factors involved in T-cell signaling/activation, such as NFAT, NF-KB and AP-1. We further characterized the role of p30(II) in regulation of viral gene expression, by identifying motifs critical in binding p300 and regulating TRE-mediated transcription. By analysing the amino acid domain of p30(II) critical for repressing LTR-mediated transcription, we identified a lysine residue at amino acid 106 of p30(II), that is critical for repressing TRE-mediated transcription. Additionally, we found that p300 reverses p30(II)-dependent repression of TRE-mediated transcription, in a dose-responsive manner. Our data confirms the role of p30(II) as a regulator of viral gene transcription, in association with p300. However, unlike wildtype p300, p300 HAT mutants only partially rescue p30(II)-mediated LTR repression. Additionally, (open full item for complete abstract)

Committee: Michael Lairmore (Advisor) Subjects: Biology, Molecular

Master of Science (MS), Bowling Green State University, 2021, Biological Sciences

Pharmaceutical companies have slowed the discovery and development of antibiotics due to low-profit margins. Therefore, antibiotic discovery is at an all-time low, and pathogens have evolved resistance to all currently available drugs. As a result, multi-drug resistant (MDR) bacterial infections are becoming more difficult to treat, especially in individuals at a high risk for infection such as cystic fibrosis (CF) patients. CF is a genetically inherited disease that inhibits or decreases chloride ion transport across epithelial cell membranes, resulting increased mucus viscosity, impairing normal clearance in the lungs. This environment is ideal for bacterial colonization and leads to a chronic lung infection. A major pathogen that colonizes the CF lung over time is Pseudomonas aeruginosa. A promising alternative treatment against MDR P. aeruginosa is bacteriophage therapy which has several advantages compared to antibiotics. First, phage therapy exhibits minimal side effects because phage are highly host-specific and do not inhibit other bacteria that are part of the human microbiome. Second, phage replicate itself exponentially when killing its host; and third, phage can be applied directly to the site of infection. However, like antibiotics, bacteria can evolve resistance to phage. To circumvent the problem of phage and drug resistance, trade-off effects may promote opportunities against both entities that may be exploited to treat MDR infections. I hypothesize that the effectiveness of antibiotics can be restored after selective pressure from bacteriophage. To test this hypothesis, MDR P. aeruginosa strains were exposed to phage in trade-off experiments, and results showed that the evolved phage resistant P. aeruginosa strain became antibiotic susceptible. In one trade-off experiment, a temperate phage recombined in the P. aeruginosa pathogen at a location downstream of a multidrug resistance efflux pump that may directly affect antibiotic susceptibility. In an (open full item for complete abstract)

Committee: Hans Wildschutte Ph.D (Advisor); George Bullerjahn Ph.D (Committee Member); Ray Larsen Ph.D (Committee Member) Subjects: Bioinformatics; Biology; Biomedical Research; Microbiology; Molecular Biology

Doctor of Philosophy, Case Western Reserve University, 2021, Systems Biology and Bioinformatics

Transcription factors(TFs) are specialized DNA binding proteins, that regulate target gene (TG) expression by driving the process of transcription. Disruption of TF binding sites can cause significant changes in TG expression, which has been shown to be associated with several diseases. Moreover, besides binding of TFs, which is a local/cis regulatory mechanism, trans-acting mechanisms such as cooperativity among different combinations of TFs and co-regulation of multiple TGs by the same set of TFs can also influence TG expression. Integrative approaches that can incorporate information from these mechanisms, obtained from different data sources, are needed to comprehend TF based TG expression regulation on a systems-wide level. Furthermore, there is also a need to integrate this regulatory information in tests for statistical associations of genetic variants with complex disease traits to unravel mechanisms responsible for causing these diseases, while also discovering novel risk TGs. In this dissertation, I develop an integrative gene regulatory network based approach utilizing information from different cis and trans regulatory mechanisms to model TG expression using machine learning algorithms. Furthermore, I use these models to calculate effect estimates of individual TFs as well as of combinations of TFs forming TF regulatory modules. Lastly, I build neural networks to quantify influence of non-coding variants on TF binding and integrate them with effect estimates of the TFs in order to derive their impact on TG regulation. I utilize these aggregated scores, as weights for common variants (allele frequency > 5%), to build TG expression prediction models based on individual level genotype information to perform transcriptome wide association study(TWAS) within my novel framework TFXcan. I show that such models are more accurate compared to state-of-the-art TWAS models using broad epigenetic priors as variant weights. Furthermore, I describe a novel weighted k (open full item for complete abstract)

Committee: William Bush (Advisor); David Lodowski (Committee Chair); Mark Cameron (Committee Member); Rong Xu (Committee Member) Subjects: Bioinformatics; Genetics; Molecular Biology

Doctor of Philosophy (PhD), Ohio University, 2020, Plant Biology (Arts and Sciences)

Flowering plants produce two different kinds of flowers: chasmogamous, open flowers that promote cross-pollination, and cleistogamous, small mechanically sealed flowers that force self-pollination. Typically, plants produce either chasmogamous or cleistogamous flowers. However, some species contain a chasmogamous/cleistogamous mixed breeding system and develop both flowers at different locations along the plant or at different times of the season. The temporal separation of chasmogamous and cleistogamous flowers is hypothesized to be influenced by the environment. However, the specific environmental cues and genetic mechanisms that signal for the induction and differentiation of each flower type remain largely unknown. To analyze the environmental variables driving separate, seasonal development of chasmogamous and cleistogamous flowers in Viola pubescens, bud counts and measurements of light quantity, canopy cover, photoperiod, temperature, soil moisture, and soil pH were collected over a native population in 2016 and 2017. Statistical analyses confirm that the two flowers develop in response to differing environmental factors and distinguish which factors are significantly associated with chasmogamous versus cleistogamous bud presence, the number of buds produced, and specific threshold values at which each bud type is most likely to develop. To enable genetic investigations, the V. pubescens draft genome was assembled using DNA and RNA sequencing data from eight diverse tissues, which together generated 38,081 gene models including 81 novel cyclotide genes. Gene expression was visualized for each tissue and highlighted transcriptional networks specific to chasmogamous and cleistogamous flowers. Functional annotation of the flower-specific genes indicate that chasmogamous flowers are enriched with genes involved in environmental sensing and repressing precocious development, while cleistogamous flowers contain up-regulated genes involved in DNA topology and microR (open full item for complete abstract)

Committee: Harvey Ballard Jr. (Advisor); Sarah Wyatt (Advisor); David Rosenthal (Committee Member); Ronan Carroll (Committee Chair) Subjects: Bioinformatics; Biology; Developmental Biology; Ecology; Environmental Science; Genetics; Plant Biology; Plant Sciences

Doctor of Philosophy, The Ohio State University, 2017, Computer Science and Engineering

Understanding the inherent genomic characteristic in the developing brain is a critical topic in developing neuroscience and translational bioinformatics. Since the brain is one of the longest and most complex developing structure, the exploration of the comprehensive patterns that describe the changes during growth poses significant challenges. With advances in data generation technologies, the Allen Developing Mouse Brain Atlas (ADMBA) and Allen Developing Human Brain Atlas (ADHBA) projects became an invaluable resource for neuroscientists and developmental biologists for exploring interesting spatial and temporal patterns of gene expression. However, given the extremely large amounts of data, it is desirable to apply visualization techniques to their access, analysis, and interpretation. This dissertation proposes an extensible visual analytics framework for the spatiotemporal pattern exploration in the developing mouse and human brain. Targeting on the ADMBA and ADHBA data, I developed three visual analytics components: the spatial pattern exploration, the region-based temporal pattern exploration, and the integrative gene gradient-based spatiotemporal pattern exploration. The spatial pattern exploration component, HOS-Tree system, uses a tree-layout visualization to present the structural hierarchy and uses colors to indicate the developing orientations. Also, the region-based temporal pattern exploration component uses data-mining approaches to provide interactive pattern presentation among genes and structures. In addition, we use the gradient of gene expression to define the spatiotemporal genomic characteristic, and also design a 3-D visualization component to provide the exploration of the spatiotemporal patterns. For each visualization component, I investigate the visual analytics result by seeking the biological interpretation of the explored patterns. The investigation shows that several explorations are well-interpreted by development ground truth. T (open full item for complete abstract)

Committee: Raghu Machiraju (Committee Chair); Kun Huang (Committee Co-Chair); Christopher Bartlett (Committee Member) Subjects: Bioinformatics; Computer Science

Doctor of Philosophy, Case Western Reserve University, 2016, Molecular Medicine

Elucidating the molecular determinants of aggressive prostate cancer has major implications for developing molecular diagnostics that, in conjunction with clinicopatholgic variables, can be used to risk-stratify patients more accurately. We hypothesized that DNA methylation can distinguish indolent from aggressive prostate cancer through aberrant gene regulation. This thesis presents the novel bioinformatics tools needed to pursue this hypothesis, a methylome analysis in aggressive prostate cancer, evidence for genomic context-specific functional interpretations of DNA methylation, data from a model system regarding a role for DNA methylation in regulating alternative cleavage and polyadenylation (APA), and integration of differential methylation data with a machine learning predictor of prostate cancer histology. We found focal DNA hypermethylation can distinguish indolent, low Gleason grade prostate cancer from aggressive, high grade disease using the genome-wide MBD-isolated genome sequencing technique. Notably, high grade DMRs occurred not only in the promoters of genes, but also in gene bodies and intergenic regions highly enriched for DNaseI hypersensitivity, annotated enhancers, and DNA-protein binding sites. DNA methylation also correlates with reduced gene expression for genes where low expression associates with poorer outcomes. By analyzing genic DMRs in a pathway network, we detected common cancer-related signatures between shared and high grade DMR genes. We further extended our analysis to consider connections with alternative transcript usage. Using RNA-seq data from The Cancer Genome Atlas (TCGA), we correlated DNA methylation changes with alternative exon usage in aggressive prostate cancer. Following an observation that prostate cancer DMRs show enrichment in gene 3' ends, we used a model system with genomic depletion of DNA methylation to consider if DNA methylation can regulate alternative cleavage and polyadenylation (APA). We related candidate (open full item for complete abstract)

Committee: Angela Ting PhD (Advisor); Thomas LaFramboise PhD (Committee Chair); Eric Klein MD (Committee Member); Jonathan Smith PhD (Committee Member); Alex Almasan PhD (Committee Member) Subjects: Bioinformatics; Biomedical Research; Genetics; Molecular Biology

Doctor of Philosophy, Case Western Reserve University, 2014, Molecular Medicine

Atrial Fibrillation (AF) is the most common sustained arrhythmia, and is associated with an increased risk of mortality, morbidity and stroke. Genome wide association studies (GWAS) have identified that the single nucleotide polymorphisms (SNPs) most strongly associated with AF are located on chromosome 4q25 in an intergenic region that is closest to the PITX2 gene. The PITX2c isoform expressed specifically in the left atria plays a role in cardiac left/right asymmetry during development and hemizygous knockout mice are susceptible to pacing induced arrhythmia. The Cleveland Clinic Lone AF GWAS identified four independent AF risk SNPs at the chromosome 4q25 locus, with the most significant SNP identified located ~171kb distal to PITX2c. In addition to PITX2c, our group identified an uncharacterized PITX2 Adjacent long intergenic Noncoding RNA (PANCR), ~2 kb proximal to PITX2. We hypothesized that the AF associated SNPs located on chromosome 4q25 might directly affect expression of PITX2c and/or PANCR. We determined that PITX2c and PANCR levels were positively correlated with each other in 223 left atrial appendages. Expression of both genes was examined in a panel of 33 human tissues, and both were highly expressed in left atria and eye. Using left atrial appendages surgically obtained from subjects of European ancestry, we obtained SNP genotypes via microarray and measured PITX2c and PANCR expression via quantitative RT-PCR (qRT-PCR). We found that the AF risk SNPs were not associated with PITX2c or PANCR expression. During differentiation of H9 human embryonic stem cells into cardiomyocytes, both transcripts were induced early prior to expression of cardiac troponin, suggesting they are coordinately expressed. Knock-down of PANCR in differentiated H9 cardiomyocytes led to decreased expression of itself and of PITX2c. RNAseq after knock-down of PITX2 and PANCR independently and simultaneously revealed a large number of genes that were differentially alt (open full item for complete abstract)

Committee: Jonathan Smith Ph.D. (Advisor); Peter Scacheri Ph.D. (Committee Chair); Mina Chung M.D. (Committee Member); Micheala Aldred Ph.D. (Committee Member); Christine Moravec Ph.D. (Committee Member); David Van Wagoner Ph.D. (Committee Member) Subjects: Cellular Biology; Genetics; Molecular Biology

Master of Science, University of Akron, 2013, Computer Science

DNA Microarray technology has provided a very convenient way to concurrently investigate the expression levels of thousands of genes in a collection of related samples during different biological processes. Researchers from different disciplines such as computer science and biology have found it very much interesting and meaningful to group genes based on the similarity of their expression patterns. Different clustering algorithms such as hierarchical clustering, k-means clustering, self-organizing maps have been applied to group of genes with similar expression patterns. However each of these traditional clustering algorithms suffers from different limitations. Beside these clustering algorithms, there are some other algorithms to group similar items together. Ford Fulkerson algorithm which is based on maximum flow – minimum cut approach is one of them and it is widely used for community discovery in web graphs. In this research work, we aimed to group genes with similar expression pattern using two different approaches: one is k-means clustering combined with hierarchical clustering and another is maximum flow – minimum cut approach in association with Dijkstra's algorithm to select source and sink node. We use a publicly available microarray data on Adenocarcinoma which is the most frequent type of non-small-cell cancers. This dataset is available in the Gene Expression Omnibus which is a public functional genomics data repository. This dataset contains samples of five different groups: normal tissue, tissues with EGFR mutation, tissues with KRAS mutation, tissues with EML4-ALK fusion and tissues with EGFR, KRAS, EML4-ALK negative cases. We investigate a number of representative genes from the group of normal tissue and from the group of KRAS mutation tissues which is also termed as KRAS positive groups in this study. We clustered the genes for both of these groups. Finally we used Gene Ontology database to find the change in the enrichment of molecular functi (open full item for complete abstract)

Committee: Zhong-Hui Duan Dr. (Advisor); Yingcai Xiao Dr. (Committee Member); Kathy Liszka Dr. (Committee Member) Subjects: Computer Science

Doctor of Philosophy (PhD), Wright State University, 2010, Biomedical Sciences PhD

A comprehensive understanding of the molecular signaling pathways that regulate cell growth and proliferation is essential in the realization of new therapeutic options to facilitate early detection and eradication of malignancy. Understanding the transcriptional regulation of the YPEL3 and FHIT genes forms the basis of this dissertation. YPEL3, or Yippee-like 3, is a newly identified p53 target gene that inhibits tumor cell growth and is thus itself, a novel tumor suppressor gene. FHIT, or Fragile histidine triad, is a well known tumor suppressor gene and is regulated at the transcriptional level by another growth inhibitory protein, FOXO3a, a Forkhead box transcription factor. Our laboratory has determined that YPEL3 is a direct transcriptional target of the tumor suppressor gene p53. The first section of this dissertation provides significant experimental evidence to validate this observation. Briefly, YPEL3 was shown to be upregulated downstream of p53 protein stabilization in a microarray screen that explored global gene expression modulation after RNAi-mediated reduction of p53's negative regulators, Hdm2 and HdmX. Genotoxic stress induced by treatment with DNA-damaging agents resulted in stabilization of p53 protein along with elevation of YPEL3 transcript and protein levels. Moreover, there exists a cis-acting p53 response element within the YPEL3 promoter that is bound by p53 in response to this stress. YPEL3 also elicits growth inhibition and decreases in colony formation when expressed in tumor cells. It is apparent that cells which express exogenous YPEL3 are forced into a permanent cell cycle arrest, termed “premature senescence.” Similar growth suppressive phenotypes are typical of many other known p53 target genes. However, most of these genes are associated with the regulation of transient inhibition of cell division or the induction of apoptosis. YPEL3 is unique in its ability to trigger premature cellular senescence. Indeed, YPEL3 stands out among (open full item for complete abstract)

Committee: Steven J. Berberich PhD (Advisor); Robert W. Putnam PhD (Committee Chair); Gerald M. Alter PhD (Committee Member); Michael Leffak PhD (Committee Chair); Paula M. Termuhlen MD (Committee Member) Subjects: Molecular Biology

Doctor of Philosophy, The Ohio State University, 2004, Veterinary Biosciences

Retrovirus-derived vectors provide efficient means of gene-transfer and are potential tools for therapeutic intervention against congenital or inherited disorders by in-utero gene therapy and for gene-transfer into dividing cells both in vitro and in vivo. To improve biodistribution of retroviral gene-transfer vectors and to obtain efficient gene-transfer towards hematopoietic stem cells via pre-natal administration, we performed in-utero administration of retroviral vectors pseudotyped with different envelope glycoproteins. Relatively high gene-transfer was obtained with significant differences in biodistribution and gene-transfer efficiency using various pseudotypes. There was significant reduction in the percentage of colonies containing transgene, 3 months postnatally. Next, we used HIV-derived lentiviral vectors to investigate the role of human T-lymphotropic virus type-1 (HTLV-1) accessory protein p12(I) in viral pathogenesis. HTLV-1, a deltaretrovirus cause adult T-cell leukemia/lymphoma, and other lymphocyte-mediated disorders. HTLV-1 provirus encodes various regulatory and accessory genes, including ORF-I encoded p12(I). We have demonstrated that this endoplasmic reticulum-localizing protein, increases intracellular calcium, activates NFAT-mediated transcription and is critical for infectivity in vivo and in vitro. Herein, using microarrays, we tested the role of p12(I) in regulating cellular gene expression in T-lymphocytes and found that p12(I) altered the expression of several calcium-regulated genes and genes associated with T-cell signaling, cell-proliferation and apoptosis. Furthermore, p12(I) upregulated the levels of p300 to transcriptionally significant levels, in T-lymphocytes. Next, we further characterized the mechanism of p300 upregulation by p12(I) and demonstrate that it is calcium-dependent, but calcineurin-independent. Sustained low-magnitude calcium release results in increased p300 RNA, protein and p300-mediated transcriptional activity i (open full item for complete abstract)

Committee: Michael Lairmore (Advisor) Subjects: Biology, Molecular

Doctor of Philosophy, The Ohio State University, 2003, Neuroscience

Our concept of multiple sclerosis (MS) has expanded from an inflammatory demyelinating disease of the central nervous system (CNS) to one incorporating the concept of MS having an inflammatory mediated neurodegenerative component causing extensive neuronal damage resulting in permanent neurological disability. The majority (80-90%) of typical MS patients have a relapsing-remitting (RR) disease course. Early in the disease, RR-MS patients often benefit from anti-inflammatory therapeutics and recover neurological function with resolution of inflammation and edema. Eventually, many patients enter the secondary progressive phase of the disease characterized by increasing neurological disability, minimal recovery and poor efficacy of anti-inflammatory therapeutics. Why is the progression of RR-MS biphasic? Demyelination and inflammation certainly cause reversible neurological disability and may cause permanent neurological disability in MS. However, neurological decline doesn't always associate with inflammatory lesions in MS and recent evidence indicates chronically demyelinated axons are capable of conducting signals suggesting irreversible neurological disability is due to additional mechanisms. We identified significant amounts of axonal transection in WM lesions from MS patients with short disease duration, indicating axonal transection occurs from disease onset. Additionally, we identified numerous cortical lesions in MS brains containing extensive neuronal pathology characterized by transected axons, transected dendrites, and apoptotic neurons. Widespread cortical pathology in MS brains identifies another component in the physiopathogenesis of MS and further implicates neuronal pathology in the development of irreversible neurological disabilities. We along with others suggest recovery early in the disease from neurological disability is due to resolution of inflammation and edema, remyelination, and the capacity of the CNS to compensate for neuronal injury. Event (open full item for complete abstract)

Committee: Michael Beattie (Advisor) Subjects:

Doctor of Philosophy, Case Western Reserve University, 2008, Statistics

Statistical data mining is one of the most active research areas. In this thesis we develop two new data mining procedures and explore their applications to genetic data. The first procedure is called PfCluster - Profile Cluster Analysis. It is a clustering method designed for profiled genetic data. The PfCluster is efficient and flexible in uncovering clusters determined by a new class of biologically meaningful distance metrics. A new internal quality measure of clusters, coherence index, is developed to find coherent clusters. An efficient mechanism for choosing the threshold of coherent clusters is also derived and implemented. The threshold is based on the first and second order approximations to the true threshold under a null distribution for parallel clusters. The PfCluster has been applied to simulated data and two real data examples: a biomarker LOH dataset and a microarray gene expression dataset. PfCluster is competitive to the correlation-based clustering procedures. The second procedure is called RPselection - Resampling based partitioning selection. It is a feature selection algorithm designed for microarray studies. It selects a subset of genes that maximizes a fitness score. The fitness score measures the relevance between the partition labels from a clustering result and an external class label derived from the clinical outcomes. The score is computed using a resampling procedure. The RPselection algorithm has been applied to simulated data and a real uveal melanoma gene expression data. RPselection outperforms gene-by-gene test-based feature selection procedures. Software development is an integral part of modern statistical research. Two software packages, pfclust and rpselect, are developed in this thesis based on our PfCluster method and RPselection algorithm. Packages pfclust and rpselect are implemented based on R object-oriented programming framework, and they can be easily customized and extended by users. The ideas in our two procedures ca (open full item for complete abstract)

Committee: Jiayang Sun (Advisor) Subjects: Statistics

Master of Science, University of Akron, 2011, Computer Science

A tremendous increase in genomic data has encouraged biologists to turn to bioinformatics in order to assist in its interpretation and processing. One of the present challenges that need to be overcome in order to understand this data more completely is the development of a reliable method to accurately predict the function of a protein from its genomic information. This study focuses on developing an effective algorithm for protein function prediction. The algorithm is based on proteins that have similar expression patterns. The similarity of the expression data is determined using a novel measure, the slope matrix. The slope matrix introduces a normalized method for the comparison of expression levels throughout a proteome. The algorithm is tested using real microarray gene expression data. Their functions are characterized using gene ontology annotations. The results of the case study indicate the protein function prediction algorithm developed is comparable to the prediction algorithms that are based on the annotations of homologous proteins.

Committee: Zhong-Hui Duan Dr. (Advisor); Chien-Chung Chan Dr. (Committee Member); Yingcai Xiao Dr. (Committee Member) Subjects: Bioinformatics; Computer Science

Bachelor of Science, Wittenberg University, 2024, Biochemistry/Molecular Biology

Obsessive compulsive disorder (OCD) is a neuropsychiatric disorder in which individuals can suffer from obsession, compulsions, and uncontrollable thoughts. Current treatments for OCD are not effective for all individuals, but an over-the-counter drug, N-acetyl cysteine (NAC), has shown promising therapeutic effects for those affiliated with OCD. In a past behavioral study, mice were separated into groups and given NAC for one or three weeks and then given RU24969, which is a 5-HT1B serotonin receptor agonist that induces perseverative behavior, a common characteristic of OCD in mice. The mouse brains were then homogenized, mRNA was extracted, and cDNA was created. The Quantitative Polymerase Chain Reaction (qPCR) analysis was used to measure gene expression in six chosen genes: CNTN4, TNR, GRIK3, GRIA2, SLC6A4, and ADRA2A and two reference genes. Gene expression was measured in seven different treatment groups with varying conditions including normal mice, those with OCD like symptoms, and those treated with NAC. These genes are involved in the nervous system in some way, which could mean there is a link between their activity and OCD. The trend in the data for these genes showed significant increases in gene expression in the 1-week NAC + water treatment and decreases in the 3-week NAC + RU treatment.

Committee: Michelle McWhorter (Advisor); Gwynne Davis (Committee Member); Cathy Pederson (Committee Member) Subjects: Mental Health; Neurobiology; Neurology; Neurosciences; Psychology

Doctor of Philosophy, Case Western Reserve University, 2025, Biochemistry

Appropriate levels of gene expression are required for cellular fitness and survival. Messenger RNAs (mRNAs) play a central role in gene expression and thus must be tightly regulated. Despite most mRNAs undergoing degradation through the same decay pathway, mRNA half-lives are highly disparate and can vary by orders of magnitude. Some transcript-specific features, such as stable structures in the 5' UTR or regulatory sequences in the 3' UTR, can affect mRNA expression. However, these elements are not sufficient to explain the variation in mRNA stability observed transcriptome-wide. Our lab originally identified codon optimality as a global determinant of mRNA stability. Codon optimality refers to how efficiently each codon is translated and is dependent on the relative availability of charged cognate tRNAs. The most likely explanation for the observed link between codon optimality and mRNA stability is that ribosome decoding rates influence mRNA stability. However, a transcriptome-wide link between the contribution of codons to translation rate and to mRNA stability has not been previously established. Using ribosome profiling in Saccharomyces cerevisiae, we demonstrate that codon-level and transcript-level elongation rates indeed globally correlate with mRNA stability. Chemical modification of mRNA nucleosides provides an additional layer of regulation. Prior to our work, N4-acetylcytidine (ac4C) was an understudied modification solely found to exist in polyA-selected mRNA by HPLC/MS-MS. Through a collaborative effort, we identified more than 2,000 human mRNAs that are modified with ac4C, and our studies revealed novel roles for ac4C in promoting mRNA stability and translation. We also find that the acetylating enzyme N-acetyltransferase 10 (NAT10) associates with translating ribosomes, suggesting that it may monitor translation and co-translationally acetylate mRNA. Furthermore, ac4C displayed stronger base-pairing interactions with guanosine (open full item for complete abstract)

Committee: Jeff Coller (Advisor); Hung-Ying Kao (Advisor); Jonatha Gott (Committee Chair); Ashleigh Schaffer (Committee Member); Xiao Li (Committee Member) Subjects: Biochemistry; Molecular Biology

Master of Science, The Ohio State University, 2008, Graduate School

Committee: Not Provided (Other) Subjects:

Master of Science, The Ohio State University, 2007, Graduate School

Committee: Not Provided (Other) Subjects: