Master of Science in Engineering, Youngstown State University, 2024, Department of Civil/Environmental and Chemical Engineering

Wearable biosensors have become a valuable tool for their promising applications in personalized medicine. Cortisol is a biomarker for various diseases and plays a key role in metabolism, blood pressure regulation, and glucose levels. In this study, we fabricated an interdigitated laser-induced graphene (LIG) biosensor for the non-faradaic impedimetric detection of cortisol in sweat. A direct laser writing technique was used to produce the LIG. Gold nanoparticles (AuNPs) were electrochemically deposited onto the surface to enhance impedance response. A Self-Assembled Monolayer (SAM) was formed with on the AuNPs via 3-Mercaptopropionic acid (MPA) thiol chemistry. The carboxylic acid (-COOH) groups of the MPA were activated using EDC/NHS chemistry. Following activation, anti-cortisol antibodies were immobilized on the surface. Lastly, the LIG was incubated in the blocking agent bovine serum albumin (BSA) to avoid unwanted detection. Surface characterization of the LIG was performed at each step of modification by Electrochemical impedance spectroscopy (EIS) in a phosphate buffered saline (PBS) solution containing a 5 mM Fe(CN)3-/4- (1:1) redox couple. Further characterization of the modified LIG electrode was achieved through Fourier transform infrared (FT-IR), surfaced-enhanced Raman spectroscopy (SERS), and X-ray diffraction (XRD). The detection experiment using EIS was conducted in increasing concentrations of cortisol (0.1 pM-100 nM) in PBS. The ZMod decreased logarithmically (R2=0.97) with a 0.0085 nM limit of detection. Reproducibility was examined by percent change of ZMod at 100 nM and a 5.93%RSD (n=5) was observed. Additional analysis of sensor specificity and interference studies showed no substantial effect on detection. This research establishes the feasibility of using the gold nanoparticle decorated LIG electrode for flexible, wearable cortisol sensing devices, which would pave the way towards an end-user easy-to-handle biosensors as point-of-care diagno (open full item for complete abstract)

Committee: Byung-Wook Park PhD (Advisor); Frank Li PhD (Committee Member); Jonathan Caguiat PhD (Committee Member) Subjects: Biochemistry; Chemical Engineering; Chemistry; Engineering

Doctor of Philosophy, The Ohio State University, 2023, Linguistics

Given a pair of sentences, a premise and a hypothesis, the task of natural language inference (NLI) consists of identifying whether the hypothesis is true (Entailment), false (Contradiction), or neither (Neutral), assuming that the premise is true. NLI is arguably one of the most important tasks for natural language understanding. Datasets have been collected in which pairs of sentences are annotated by multiple annotators with one of the three labels. However, it has been shown that annotation disagreement, or human label variation (Plank, 2022), is prevalent and systematic for NLI – human annotators sometimes do not give the same label for the same pair of sentences (Pavlick and Kwiatkowski, 2019, i.a.). Label variation questions the widespread assumption in natural language processing that each item has a single ground truth label and casts doubt on the validity of measuring models' ability to produce such ground truth labels. In this dissertation, I investigate the question of why there is label variation in NLI and how to build models to capture it. First I analyze the reasons for label variation from the perspective of linguists, by developing a taxonomy of reasons for label variation. I found that NLI label variation can arise out of a wide range of reasons: some are due to uncertainty in the sentence meaning, while others are inherent to the NLI task definition. However, it is unclear how well the perspective of linguists reflect that of linguistically-uninformed annotators. Therefore, I collect annotators' explanations for the NLI labels they chose, creating the LiveNLI dataset containing ecologically valid explanations. I found that the annotators' reasons for label variation are similar to the taxonomy across the board, but some other reasons also emerged. Explanations also reveal that there exists within-label variation: annotators can choose the same label for different reasons. There is thus a wide range of variation that NLI models should capture. (open full item for complete abstract)

Committee: Marie-Catherine de Marneffe (Advisor); Michael White (Committee Member); Chenhao Tan (Committee Member); Micha Elsner (Committee Member) Subjects: Computer Science; Linguistics

Master of Science in Cyber Security (M.S.C.S.), Wright State University, 2023, Computer Science

Malware detection is a critical task in ensuring the security of computer systems. Due to a surge in malware and the malware program sophistication, machine learning methods have been developed to perform such a task with great success. To further learn structural semantics, Graph Neural Networks abbreviated as GNNs have emerged as a recent practice for malware detection by modeling the relationships between various components of a program as a graph, which deliver promising detection performance improvement. However, this line of research attends to individual programs while overlooking program interactions; also, these GNNs tend to perform feature aggregation from neighbors without considering any label information and significantly suffer from over-smoothing on node presentations. To address these issues, this thesis constructs a graph over program collection to capture the program relations and designs two enhanced graph convolutional network (GCN)architectures for malware detection. More specifically, the first proposed GCN model in-corporates label propagation into GCN to take advantage of label information for facilitating neighborhood aggregation, which is used to propagate labels from the labeled nodes to the unlabeled nodes; the second proposed GCN model introduces residual connections between the original node features and the node representations produced by GCN layer to enhance the flow of information through the network and address over-smoothing is-sue. The experimental results after substantial experiments show that the proposed models outperform the baseline GCN and classic machine learning methods for malware detection, which confirm their effectiveness in program representation learning using either label propagation or residual connections and malware detection using program graph. Furthermore, these models can be used for other graph-based tasks other than malware detection, demonstrating their versatility and promise.

Committee: Lingwei Chen Ph.D. (Advisor); Meilin Liu Ph.D. (Committee Member); Junjie Zhang Ph.D. (Committee Member) Subjects: Computer Science; Information Science

Master of Sciences, Case Western Reserve University, 2023, EECS - Computer and Information Sciences

Label propagation is frequently encountered in machine learning and data mining applications on graphs, either as a standalone problem or as part of node classification. Many label propagation algorithms utilize random walks (or network propagation), which provide limited ability to take into account negatively-labeled nodes (i.e., nodes that are known to be not associated with the label of interest). Specialized algorithms to incorporate negatively labeled samples generally focus on learning or readjusting the edge weights to drive walks away from negatively-labeled nodes and toward positively-labeled nodes. This approach has several disadvantages, as it increases the number of parameters to be learned, and does not necessarily avoid regions of the network that are rich in negatively-labeled nodes. We reformulate random walk with restarts and network propagation to enable “variable restarts”, that is the increased likelihood of restarting at a positively-labeled node when a negatively-labeled node is encountered. Based on this reformulation, we develop CusTaRd, an algorithm that effectively combines variable restart probabilities and edge re-weighting to avoid negatively-labeled nodes. To assess the performance of CusTaRd, we perform comprehensive experiments on four network datasets commonly used in benchmarking label propagation and node classification algorithms. Our results show that CusTaRd consistently outperforms competing algorithms that learn/readjust edge weights, when negatives are available in the close neighborhood of positives.

Committee: Mehmet Koyutürk (Advisor); An Wang (Committee Member); Yinghui Wu (Committee Member) Subjects: Computer Science

PhD, University of Cincinnati, 2021, Engineering and Applied Science: Electrical Engineering

Exosomes are nanoscale membrane vesicles (40-150 nm) produced by almost all kinds of cells, and they carry a wide variety of functional proteins and nucleic acids (particularly messenger RNAs and micro RNAs), which can reflect the status of their originating cells. It has been demonstrated that exosomes act as vehicles for molecular cargos in intercellular communication and thus have been considered as circulating biomarkers for early diagnosis and therapeutics in liquid biopsy. Beyond biomarker applications, exosomes can be used as drug-delivery vehicles with minimal immune response for targeted therapy in personalized medicine. Despite these remarkable attributes, the rapid isolation and non-invasive detection of exosomes have been a challenging task due to their small size and complex nature. The dissertation focuses on the development of a rapid, cost-effective, simple, and label-free method for exosome isolation and characterization. First, we studied the physical principle of particle trapping with a nano/micro-pipette dielectrophoretic (DEP) device. As a model system, carboxylic acid polystyrene (COOH-PS) beads have been used to comprehensively study the micro-/nano-pipette system's electrokinetic (EK) forces. The correlation between the induced EK forces and the number of trapped particles has been systematically investigated by numerical modeling and experimental observations. Next, we demonstrated the capability of the glass pipette iDEP device to isolate exosomes from conditioned cell culture media and undiluted biofluids, such as plasma, serum, and saliva from healthy donors with minimal sample preparation and high yield. The iDEP device has been demonstrated to be able to extract exosomes from 200 µL sample volumes within 20 minutes. In addition, we developed a proof-of-concept EIS microchip comprised of triangular posts on a substrate for entrapment of nanoparticles using the DEP force. The entrapped particles could be further analyzed based (open full item for complete abstract)

Committee: Leyla Esfandiari Ph.D. (Committee Chair); Chong Ahn Ph.D. (Committee Member); Scott Langevin (Committee Member); Rashmi Jha Ph.D. (Committee Member); Greg Harris Ph.D. (Committee Member) Subjects: Electrical Engineering

Doctor of Philosophy (PhD), Wright State University, 2021, Biomedical Sciences PhD

Computational models may assist in identification and prioritization of large chemical libraries. Recent experimental and data curation efforts, such as from the Tox21 consortium, have contributed towards toxicological datasets of increasing numbers of chemicals and toxicity endpoints, creating a golden opportunity for the exploration of multi-label learning and deep learning approaches in this thesis. Multi-label classification (MLC) methods may improve model predictivity by accounting for label dependence. However, current measures of label dependence, such as correlation coefficient, are inappropriate for datasets with extreme class imbalance, often seen in toxicological datasets. In this thesis, we propose a novel label dependence measure that directly models the conditional probability of a label-pair and displays greater sensitivity than correlation coefficient for labels with low prior probabilities. MLC models using data-driven label partitioning based on this measure was generally non-inferior to MLC models using random label partitioning. Marginal improvements in model predictivity have prompted toxicology modelers to shy away from deep learning and resort to ‘simpler' models, such as k-nearest neighbors, for its greater explainability. Given the prevalence of local, linear quantitative structure-activity relationship (QSAR) models in computational toxicology, we hypothesize that toxicological datasets have locally-linear data structures, resulting in heterogeneous classification spaces that challenges the basic assumptions of most machine learning algorithms. We propose the locality-sensitive deep learner, a modification of deep neural networks which uses attention mechanism to learn datapoint locality. On carefully-constructed synthetic data with extremely unbalanced classes (10% active) and (60%) cluster-specific noise, the locality-sensitive deep learner with learned feature weights retained high test performance (AUC>0.9), while the feed-forward n (open full item for complete abstract)

Committee: Michael L. Raymer Ph.D. (Advisor); David R. Cool Ph.D. (Committee Member); Lynn K. Hartzler Ph.D. (Committee Member); Travis E. Doom Ph.D. (Committee Member); Courtney E.W. Sulentic Ph.D. (Committee Member) Subjects: Computer Science; Toxicology

Doctor of Philosophy, Miami University, 2021, Chemistry and Biochemistry

As sophisticated functions in nature arise from structurally complex biomacromolecules, that has inspired the design and synthesis of new abiotic higher-order analogs. Foldamers, oligomers that can fold into well-defined conformations, are excellent building blocks to make molecules with higher-order structures. ortho-Phenylene is a simple aromatic foldamer and is already used to generate macrocycles with tertiary structures. Currently, we lack control over these o-phenylene-based macrocycles. As the first step, we introduced the 19F NMR spectroscopic method to characterize o-phenylene secondary structures. The new 19F method assisted the current 1H NMR spectroscopy-based technique to identify conformers with the help of 19F GIAO isotropic shielding calculations. The major conformer was the perfectly folded geometry, and the minor conformer was the AAB misfolded geometry. The single crystal X-ray diffraction results showed fluorinated o-phenylenes have AAA geometry in solid-state. Equally importantly, the 19F labels assisted in accurate conformer population analysis. Macrocyclization of deca(o-phenylenes) gives a mixture of products ([2+2] and [3+3] (o-phenylene+linker) macrocycles) and shows entropic product compete with the enthalpic product. o-Phenylenes in quasi-triangular [3+3] macrocycles are perfectly folded, and in [2+2], o-phenylenes are misfolded. Electron-withdrawing fluorine substituted deca(o-phenylenes), placed at a site remote from the imine site, exclusively gave the [3+3] macrocycles. Substituents change the folding propensity that controls the macrocycle product distribution by altering the subtle balance between entropic and enthalpic favorability. The fluorinated octa(o-phenylene) gave [2+2] macrocycles showed fluorinated o-phenylene can still misfold when the conditions are right. Further control over macrocyclization can be expected by introducing desymmetrized o-phenylenes for the macrocyclization. The o-phenylenes synthesis was desig (open full item for complete abstract)

Committee: Scott Hartley Dr./PhD (Advisor); Dominik Konkolewicz Dr./PhD (Committee Chair); Zhijiang Ye Dr./PhD (Committee Member); David Tierney Dr./PhD (Committee Member); Richard Taylor Dr./PhD (Committee Member) Subjects: Organic Chemistry

Master of Science, The Ohio State University, 2021, Animal Sciences

Annually, approximately 6.5 million companion animals enter U.S. animal shelters nationwide—3.3 million are dogs. Upon intake, shelter personnel evaluate each dog to gain information regarding behavioral and health history. Previous owner's records or shelter personnel's visual perception of the dog generates a “breed label.” However, the U.S. shelter canine population predominately consists of dogs with an unknown history and breed heritage; thus, the created “breed label” is a subjective assessment of breed type. Several studies illustrate that majority of shelter dogs are composed of more than one breed and are mislabeled compared to identified breeds from their genetic analysis. Additionally, breed labels have negatively impacted adopters' decisions due to certain breeds' stereotypes and legislation. Breed labeling is a tool animal shelters can utilize to implement specific care strategies and tailored adoption matches for the dogs in their care. However, canines are multidimensional individuals whose behavior and appearance are influenced by numerous factors; therefore, whether a breed label accurately reflects a dog's genetic makeup may not be efficient for successfully matching potential adopters with a companion. By steadily increasing the number of successful adoptions out of the shelter, fewer animals may require euthanasia. The overarching objective of this research was to assess the impact of removing dog breed labels in a local animal shelter on overall dog adoption rate and length of availability (LOA) for adoption, in addition to dog breeds commonly restricted by breed legislation or currently under county restrictions in the state of Ohio. The second objective examined the impact of visitors' demographics, perceptions, and importance of potential companion's features on adoption decisions, contingent on breed label presence or absence. A third objective utilized dog behavioral observations during a visitor interaction to identify out-of-kennel she (open full item for complete abstract)

Committee: Kelly George (Advisor); Kimberly Cole (Committee Member); Peter Neville (Committee Member) Subjects: Animal Sciences; Animals; Behavioral Psychology; Behavioral Sciences; Behaviorial Sciences; Demographics; Management; Marketing; Social Psychology; Veterinary Services; Welfare; Zoology

Master of Science, The Ohio State University, 2021, Electrical and Computer Engineering

We address the current approaches to radar signal processing, which model radar signals with several assumptions (e.g., sparse or synchronized signals) that limit their performance and use in practical applications. We propose deep learning approaches to radar signal processing which do not make such assumptions. We present well-designed deep networks, detailed training procedures, and numerical results which show our deep networks outperform current approaches. In the first part of this thesis, we consider synthetic aperture radar (SAR) image recovery and classification from sub-Nyquist samples, i.e., compressive SAR. Our approach is to first apply back-projection and then use a deep convolutional neural network (CNN) to de-alias the result. Importantly, our CNN is trained to be agnostic to the subsampling pattern. Relative to the basis pursuit (i.e., sparsity-based) approach to compressive SAR recovery, our CNN-based approach is faster and more accurate, in terms of both image recovery MSE and downstream classification accuracy, on the MSTAR dataset. In the second part of this thesis, we consider the problem of classifying multiple overlapping phase-modulated radar waveforms given raw signal data. To do this, we design a complex-valued residual deep neural network and apply data augmentations during training to make our network robust to time synchronization, pulse width, and SNR. We demonstrate that our optimized network significantly outperforms the current state-of-the-art in terms of classification accuracy, especially in the asynchronous setting.

Committee: Philip Schniter (Advisor); Emre Ertin (Committee Member) Subjects: Electrical Engineering; Engineering

PhD, University of Cincinnati, 2020, Engineering and Applied Science: Electrical Engineering

Action recognition is an important computer vision task. It focuses on identifying the behavior or the action performed by humans from images. Action recognition using various wearable sensors and videos is a well studied and well established topic. This thesis focuses on action recognition in still images, a new and challenging area of research. For example, understanding motion from static images is a difficult task as spatio-temporal features that is most commonly used for predicting actions is not available. Action recognition in still images has a variety of applications such as searching for frames in videos using action, searching a database of images using an action label, surveillance, robotic applications etc. It can also be used to give a more meaningful description of the image. The goal of this thesis is to perform action recognition in still images and infer object affordances by characterizing the interaction between the human and the object. Object affordance refers to determining the use of an object based on its physical properties. The main idea is to learn high level concepts such as action and object affordance by extracting information of the objects and their interactions in an image.

Committee: Anca Ralescu Ph.D. (Committee Chair); Kenneth Berman Ph.D. (Committee Member); Rashmi Jha Ph.D. (Committee Member); Wen-Ben Jone Ph.D. (Committee Member); Dan Ralescu Ph.D. (Committee Member) Subjects: Electrical Engineering

Doctor of Philosophy, The Ohio State University, 2018, Statistics

Finite Bayesian mixture models are often used to describe data arising from a heterogeneous population. If information is available about the differences among the groups represented by the mixture components, the model should explicitly incorporate that knowledge through informative distributional assumptions. If no such information is available, however, it is common to specify a fully exchangeable model, where, a priori, all components play the same role. A consequence of the exchangeable specification is the label-switching phenomenon, by which the mixture components can be relabeled arbitrarily without changing the posterior distribution of the model parameters. Label-switching makes direct marginal inference on features of the mixture components impossible and limits the interpretability of the model in applications where one is interested in discovering if and how the mixture components correspond to meaningful subgroups in the population. It is common practice to either prevent label-switching by imposing prior constraints on model parameters or ``undo'' label-switching with post-processing algorithms applied to the Markov chain Monte Carlo (MCMC) output from the exchangeable model. The former approach can be too restrictive and appropriate prior constraints are often difficult to specify. The latter approach does not correspond to any clearly-defined probability model, so any marginal features described using post-processed samples are consequences of the chosen algorithm, not the model itself. This work presents a model-based approach to the resolution of the label-switching phenomenon which treats a small number of observations (called the ``anchor points'') as pre-labeled. This results in a well-defined probability model (the ``anchor model'') that imposes a unique labeling on the mixture components but requires no prior knowledge of the components' relative locations or scales. Several basic properties of the anchor model are derived. T (open full item for complete abstract)

Committee: Mario Peruggia (Advisor); Steven MacEachern (Committee Member); Xinyi Xu (Committee Member); Yunzhang Zhu (Committee Member) Subjects: Statistics

Doctor of Philosophy, The Ohio State University, 2018, Food Science and Technology

Color is an important factor in consumer perception and quality of food products. Although extensively used in food industry, artificial colorants have been increasingly questioned among consumers due to potential health concerns. With current clean label trends as well as customer demands, manufactures are seeking alternatives for replacement of synthetic dyes. However, there are limited options for stable natural colors. The first overall objective of this dissertation was to explore new sources of natural colorants. We focused on petunidin-derivatives in black goji and purple potato, as they both are rich anthocyanin sources that could be promising candidates for natural food colors; the second overall objective was to explore co-pigmentation between the petunidin-derivatives and metal ions or soybean isoflavones as methods for anthocyanin color enhancement and stabilization. Currently the most prevalent anthocyanin-based pigments were cyanidin and pelargonidin derivatives (from red cabbage, black carrot, purple sweet potato, and red radish, among others). Focusing on these anthocyanidins would limited the innovation of natural colorant application. Therefore, we started from investigating the pigments in black goji, as it was reported to contain abundant petunidin-derivatives. The black goji extracts produced various vivid hues over wide ranges of pH, with red, purple, and blue colors in acidic, neutral and alkaline conditions, respectively. Cis and trans isomeric petunidin-3-p-coumaroyl-rutinoside-5-glucoside were the major pigments. The colorimetric and spectrophotometric traits of black goji anthocyanins were significantly impacted by purity, pH, acylation, and acyl moiety spatial configuration. The cis and trans isomeric petunidin-3-p-coumaroyl-rutinoside-5-glucoside displayed different color properties. They could be served in food products depending on diverse color demands. However, only 12 cis acylated anthocyanins were reported in reviewing the c (open full item for complete abstract)

Committee: M. Monica Giusti (Advisor); Luis Rodriguez-Saona (Committee Member); Christopher Simons (Committee Member); John Litchfield (Committee Member) Subjects: Food Science

Doctor of Philosophy, Miami University, 2017, Chemistry and Biochemistry

Membrane proteins play vital roles in performing biological functions and ensuring the survival of living organisms. Despite their importance, it is extremely difficult to study the structure of these proteins due to their hydrophobic nature, large sizes, and complex native environments. This dissertation highlights the development of several electron paramagnetic resonance (EPR) methods in combination with site-directed spin labeling (SDSL) to examine the local secondary structures of membrane peptides and proteins. Chapters 2 and 3 describe the successful use of Electron Spin Echo Envelope Modulation (ESEEM) spectroscopy to distinguish α-helix and 310-helix secondary structures and the use of a 13C isotopic label for this ESEEM method. Chapter 4 illustrates how a bifunctional spin label can be used to probe the topology of a membrane peptide in mechanically aligned lipid bilayers. Chapter 5 describes the use of ESEEM and Double Electron Electron (DEER) spectroscopy to probe the secondary structure of the Linker 12 region of the human vimentin protein. Finally, Chapter 6 describes the use of continuous wave (CW)-EPR to study radical formation in a pharmaceutical matrix, in a collaboration with Procter & Gamble. The development of these methods expands the application of EPR as a biophysical tool to investigate secondary structures of membrane proteins as well as its use in an industrial setting.

Committee: Gary Lorigan Dr. (Advisor); Michael Crowder Dr. (Committee Chair); Rick Page Dr. (Committee Member); David Tierney Dr. (Committee Member); Natosha Finley Dr. (Committee Member) Subjects: Biochemistry; Biophysics; Chemistry

Master of Science, The Ohio State University, 2017, Human Ecology: Family Resource Management

Facing hundreds of food choices and tons of information, time-constrained consumers always spend time and efforts to search information and seek the greatest-value option (McCall et al., 1970). Nearly 200 food decisions need to be made by each consumer per day (Wansink and Sobal, 2007), and it is hard for consumers to check all the information. When purchasing food, consumers may rely on some information resources that can be easily gathered such as health claims, front-of-package (FOP) labels, online reviews and so on to make quick decisions. There is evidence that nutrition content claims on front of packages can assist people understanding products nutrition information and choosing healthier food. Yet, it is unknown whether the information of claims affects consumers or the number of claims matters. As a result, this study was interested in investigating how the number of nutrition content claims on the front of packages influences consumers. Specifically, two primary objectives of this study are (1) to measure the effect of the number of nutrition content claims on consumers' purchase valuations of food products, regardless of the claim, (2) to determine whether or not there is a threshold point at which the number of claims does not affect product valuations. An online questionnaire was designed to test the effect of the number of nutrition content claims on consumers' purchase intentions toward food products, and their willingness to pay. In this survey, four food categories: yogurt, cereal, frozen lasagna and peanut butter were selected to test. The results indicate that for yogurt, the number of nutrition content claims influences consumers' perception and purchase valuation of food products. For cereal and peanut butter, the number of claims only has a significant effect on the purchase valuation of products. And for frozen lasagna, no significant difference of perception or purchase valuation is found between food products with a different number of n (open full item for complete abstract)

Committee: Andrew Hanks (Advisor); Jay Kandampully (Committee Member); Robert Scharff (Committee Member) Subjects: Food Science; Individual and Family Studies

Doctor of Philosophy, The Ohio State University, 2016, Biochemistry Program, Ohio State

This dissertation describes the research that was conducted on the development of label-free quantitation procedures for the identification and quantitation of proteome differences determined from shotgun proteomics experiments. Chapter 1 introduces common approaches of which their basic understanding of is imperative for all proteomic scientists. This introductory chapter also describes label-free quantitation approaches, which is built upon in following chapters. Chapter 2 outlines a novel approach to perform label-free spectral count quantitation from shotgun proteomic experiments. This approach, termed MultiSpec, utilizes open-source statistical platforms; namely edgeR, DESeq and baySeq, to statistically select protein candidates for further investigation. The results from these three statistical approaches are combined to provide a single ranked list of differentially expressed proteins. The statistical results from multiple proteomic pipelines are integrated and cross-validated by means of collapsing protein groups. Chapter 3 highlights the efficient application of negative binomial based tag-count analysis of large-scale proteomics. This chapter illustrates the efficacy of edgeR to perform spectral count quantitation across a large number of samples. Chapter 4 demonstrates the use of precursor abundance (MS1) quantitation, an alternative to spectral count quantitation, to quantitate proteome differences in chromatin-bound androgen receptor protein complexes pivotal in directing proper gene expression in the context of localized human prostate cancer. Also presented in chapter 4, precursor intensities were used to determine proteome differences between the prostate proteomes of a transgenic mouse model of prostatic intraepithelial neoplasia (PIN). In a collaborative effort, these data were overlaid with RNA sequencing and Chromatin-Immunoprecipitation sequencing data to identify a proteome set of putative androgen receptor regulated proteins.

Committee: Michael Freitas (Advisor) Subjects: Biochemistry

Doctor of Philosophy, The Ohio State University, 2016, Chemistry

Even with extensive genomic and transcriptomic characterization of tumors, the relationship of the human cancer genotype to cancer phenotype remains unclear. Proteins, however, are the immediate molecular drivers of the cancer phenotype that govern tumorigenesis or tumor recurrence. The research described here highlights work on non-small cell lung cancer tumors and cell lines. The major goals are to discover proteins exclusive to tumor recurrence and liver kinase B1 (LKB1) gene mutation, respectively. The proteins were discovered by nanoflow multidimensional liquid chromatography coupled to mass spectrometry. The goal of the research described in Chapter 2 of the dissertation focuses on establishing a mass spectrometry-based bottom-up proteomic method for protein detection in formalin-fixed paraffin embedded (FFPE) tissue specimens. Identification of protein markers for lung cancer requires tumor tissues that are usually unavailable in the fresh, frozen state. FFPE tissues, however, are produced from resected tumor material and are readily available for proteome analysis. The use of these tumor samples in mass spectrometry demands effective sample preparation and detection strategies. In the analysis, the use of an on-slide deparaffinization method and modified lysis buffer recovered the maximum amount of protein from the slide tissue specimen and reduced the sample incubation time during digestion. Fractionation of peptide digests into fifteen high pH reversed phase fractions followed by low pH reversed phase separation resulted in the highest number of protein identifications for a minimum amount of tissue protein extract when coupled to an optimized mass spectrometry method. In Chapter 3, the use of this method for the tumor protein analysis yielded over five thousand proteins per cohort. The corresponding changes at protein level were identified by comparing the proteins discovered in specimens from recurrent to those of nonrecurrent patients. Adenocarcinom (open full item for complete abstract)

Committee: Vicki Wysocki (Advisor); David Carbone (Committee Member); Susan Olesik (Committee Member); Abraham Badu-Tawiah (Committee Member) Subjects: Chemistry

Doctor of Philosophy, The Ohio State University, 2015, EDU Policy and Leadership

This qualitative study used interview and focus group data from 13 gay, lesbian, bisexual, queer (GLBQ), and other non-heterosexual students of color to add to the extant literature on the intersections of race, sexuality, gender, and other social identity categories in higher education. Using a grounded theory methodology supplemented by intersectionality as its theoretical framework, this dissertation study offers a number of findings that increase our understanding of the ways in which GLBQ college students of color understand, navigate, negotiate, and enact sexual identity label adoption and sexuality disclosure possibilities. The first set of findings explore sexual identity label adoption. In their discussion of label adoption considerations, participants describe sexual identity labels as possessing a utilitarian function; that is, operating as a tool rather than just a descriptor of their sexuality. To this end, there were five findings that emerged regarding sexual identity label adoption; collectively they include the following considerations: (a) a willingness to adopt a sexual identity label, (b) the nature of the adoption process being less than straightforward, (c) need to adopt alternate sexual identity labels to be able to share that identity, (d) the influence of sexual identity development and label adoption of one's understanding of race, and (e) association between access to diverse array of sexual identity labels and one's academic and social involvement. Findings related to sexuality disclosure primary focused on three areas: motivation for disclosure, impetus to conceal or not vocalize one's sexuality and sexual identity, and additional factors that influence disclosure. In addition to findings, implications for research, policy, and theory are considered.

Committee: Terrell Strayhorn (Advisor); Wendy Smooth (Committee Member); Shannon Winnubst (Committee Member) Subjects: African Americans; Education; Ethnic Studies; Glbt Studies; Higher Education; Higher Education Administration

MS, Kent State University, 2014, College of Education, Health and Human Services / School of Health Sciences

The purpose of this study was to compare the knowledge and utilization of nutrition labels among international versus non-international college students. It was expected that there would be a difference in knowledge of the nutrition labels between the international and non-international college students. Also, it was expected that there would be a difference in utilization of nutrition labels among international and non-international college students. An electronic questionnaire was completed by undergraduate and graduate students at Kent State University (n=176). Descriptive statistics were utilized to describe frequencies, standard deviations, and means of all participants' responses. A t test was used to compare the means of the three subscales (nutrition knowledge, nutrition label use, and attitude toward nutrition labels) among the demographic variables. A P-value was selected a priori 0.05 for significance. Correlation between age and the three scales was used to analyze the relationship between age and scores on each of the three scales. A significant difference was demonstrated in the summed total knowledge scores between non-international and International students (P=.001). This study demonstrated a lack of overall nutrition label knowledge and use among college students, suggesting nutritional-related educational strategies for college students are needed.

Committee: Karen Gordon Ph.D., R.D., L.D. (Advisor); Natalie Caine-Bish Ph.D., R.D., L.D. (Committee Member); Amy Miracle Ph.D. R.D., CSSD (Committee Member) Subjects: Food Science; Health; Health Sciences; Nutrition

MS, University of Cincinnati, 2014, Engineering and Applied Science: Computer Science

Twitter is one of the most popular social networking services. By analyzing social network data, meaningful information can be discovered, such as popular topics users are discussing and trends of important events. This thesis focuses on the problem of discovering and optimizing semantic labeling of network clusters using Twitter data sets. Specifically, we focus on prominent, individual Twitter accounts around the University of Cincinnati. With its heavily structured nature, Twitter is an appropriate environment in which to observe social interactions and determine the level of influence a given individual exerts on the people who see their content. The data sets we used in this thesis consists of a very active group of users associated with the account of President of University of Cincinnati. By applying an algorithmic design based on Order Statistics Local Optimization Method(OSLOM) algorithm, we cluster all accounts into several groupings or clusters according to their mutual relationships. We also develop a method to label accounts by analyzing tweets from selected accounts. We eliminate stop words and label accounts by word occurrence number to find out their interests and hot topics.?

Committee: Fred Annexstein Ph.D. (Committee Chair); Wen Ben Jone Ph.D. (Committee Member); Anca Ralescu Ph.D. (Committee Member) Subjects: Computer Science

Master of Arts, The Ohio State University, 2011, Arts Policy and Administration

The arts' place in the university is changing in response to the demands of the creative economy. Universities will be responsible for producing creative human capital in their graduates. The 2004 American Assembly provided campus-based practitioners with new language to pursue these goals when it introduced the “Creative Campus” terminology. This comparative case study explores the value of this naming language during policy formulation and implementation of two Creative Campus projects. Qualitative interviews, document analysis, and autoethnography are used to assess the value of a common naming strategy. A critical framework that crosses semiotics and the policy cycle is used to analyze the data from each of the cases. The findings suggest that a symbolically isomorphic naming strategy can be very effective in formulating and implementing a Creative Campus program because the name label provides cultural entrepreneurs with a tool to contextualize their work, frame the issue on the institutional agenda, define their work in juxtaposition to a prototypical schema, and gain legitimacy, understanding, consensus, and control of resources. This thesis concludes by suggesting that the shared signifier also offers an opportunity for a more formalized network of Creative Campus practitioners to learn from and engage in the labeling contests that shape the sign.

Committee: Margaret J. Wyszomirski (Advisor); Wayne Lawson P. (Committee Member) Subjects: Art Education; Arts Management; Cultural Resources Management; Language; Linguistics; Public Policy; Rhetoric