UBIR Repository (Univ. at Buffalo)
Not a member yet
5736 research outputs found
Sort by
Adaptive Radial Basis Function Methods for Initial and Boundary Value Problems
No full textPh.D.The classical linear finite difference methods for solving initial value problems, such as Euler’s method, the midpoint method and the Adams methods, are based the polynomial interpolation. The order of every classical method is fixed regardless of the smoothness of the unknown solution. The radial basis function (RBF) interpolation is one of the most often applied approaches in modern approximation theory. We consider the corresponding RBF finite difference methods derived from the RBF interpolation. The shape parameter inherent to RBF could be used to include some information about the derivatives of the solution.In Chapter 1, we start with the one-dimensional finite difference approximations in the context of the polynomial interpolation. Then the RBF interpolation follows, and the chapter ends with the RBF finite difference approximations in the RBF approach...Chapter 2 covers several multiquadric RBF (RBF-MQ) methods for initial value problems, including RBF-MQ Euler's method, RBF-MQ midpoint method, RBF-MQ Adams methods...In Chapter 3, we continue to construct several Gaussian RBF (RBF-GA) methods for initial value problems, corresponding to the RBF-MQ methods in the previous chapter...In Chapter 4 we improve the error inhibiting block one-step methods...In Chapter 5, the RBF-GA finite difference methods on the uniform grid for singular perturbation problems are derived and improve the numerical approximation...The results from Chapter 2 have been published in Journal of Scientific Computing. The results from Chapter 3 have been published in Journal of Computational and Applied Mathematics. The results from Chapter 4 have been published in Spectral and High Order Methods for Partial Differential Equations ICOSAHOM 2018: Selected Papers from the ICOSAHOM Conference, London, UK, July 9-13, 2018.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Towards Understanding the Structural Impact from Therapeutic Radiation Doses: From Mechanisms in Engineered Systems to TGF-β1
No full textPh.D.Ionizing radiation can alter protein function and damage DNA leading to cell death, which is often undesirable but can be advantageous for the treatment of cancer. How a particular protein structure is altered from radiation exposure is often unknown. X-ray crystallography relies on ionizing radiation to reveal the specific coordinates of atoms in protein molecules. Damage to the protein molecule during these experiments is one of the major limitations of the technique and the effects on structure have been well studied. This technique and others can potentially be used to understand how radiation influences functionally relevant protein dynamics. However, because crystallographic experiments require radiation doses of several orders of magnitude greater than biomedical applications it is unclear how relatable these mechanisms are. To study the effects of low doses used in biomedical applications using structural techniques, new data collection strategies and analyses were developed. First, to determine whether the most prolific crystallographic damage phenomenon, disulfide cleavage, occurs in a physiologic setting we developed an engineered system and monitored changes from exposure to low dose X-rays using small angle solution X-ray scattering (SAXS). The underlying chemistry was investigated with small molecules that scavenge different radiation products in the solvent. This methodology was applied to the transforming growth factor beta-1 (TGF-β1) system, which has been observed to undergo a structural change from radiation exposure in cells, but the mechanism is unclear. Using structural and biochemical techniques we present a mechanism of how the TGF-β1 pro-domain, latency associated peptide (LAP), undergoes a multi-domain conformational change to sequester TGF-β1 and prohibit its activity, which reveals insight into using LAP as an anti-TGF-β1 therapy.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Multi-Camera System for 3D Human Pose Reconstruction
No full textM.S.During our research on reconstructing human skeleton or even mesh using wifi signal, we need precise 3D coordinates of human joints as grand truth to train our model. Usually we can get this information via VICON system. But when we want to collect data from a different room or open-space environment, VICON is not feasible. Although we can use other means like open-pose or deeplabcut to get such kind of information, they are too computationally expensive. Further more, there is too much estimation part in their results which makes it not so reliable as grand truth for other models' training. Therefore, I want to find a reliable and non-machine-learning way to get the exact 3d coordinates of human joints. To achieve this goal, I construct a system using several cellphones or webcams to get 3D coordinates of human joints and other objects in different environments. It uses pinhole camera model and optimizer. Technically, this system entails only python and OpenCV. The Final precision of this system can reach 0.5-1.0 cm.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
A Report on WebAssembly Compiler Bugs
No full textM.S.With the increasing need for high performance of web applications, a new standard is designed to solve this problem, which is WebAssembly. WebAssembly is supported by all major browsers now. It is a new standard that defines a new type of low-level bytecode language that can run in the browsers. The language is designed to be a compact and portable binary-code format and primarily provides low-level source languages like C/C++ and Rust as a compilation target so that they can run on the web. It provides a way to run applications written in multiple languages on the web with a near-native speed. Since first announced in 2015, more than 50 compilers are supporting compiling WebAssembly as the target for now. Although the number is gratifying, the qualities of these compilers vary a lot. This directly impacts the reliability of WebAssembly related products and the whole WebAssembly ecosystem. A comprehensive understanding of the compiler bugs can be a great help for developers either to improve the quality of compilers or to use the compilers more efficiently. Especially understanding the impact bugs caused and how to reproduce the bugs can help a lot. For this intention, we conduct this study on the impacts and the size of code to reproduce the bugs.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
The Inter-Areality of Areal Features
No full textPh.D.Descriptions of twelve language areas were consulted to find a list of areal features to use. Twenty four of these features were selected: six phonological features, six morphological features, six syntactic features, and six semantic features. This set of features was used a basis for exploring whether each feature occurred in eight language areas. In each area, samples were mostly chosen based on which languages were previously mentioned as being associated with the highest number of features in that area. When possible, the samples were also balanced so that each language family was represented by a similar number of languages. The investigation of each area produced previously undescribed areal feature values. No area had more than sixteen of the features, and no area had less than twelve. The different features ranged in the number of areas in which they were found from one to eight.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
ETS1 and Tumor Heterogeneity: Uncovering Novel Roles of ETS1 within Intrinsic Subtypes of Epithelial Cancers
No full textPh.D.Large-scale sequencing of human tumors has revealed tremendous intra and inter-tumor heterogeneity. Indeed, tumor-to-tumor variations in genetic mutations, epigenetic aberrations and gene expression are recognized in almost every type of cancer. However, current therapies to treat cancer often do not take into account tumor heterogeneity. Hence, understanding the specific driving forces behind different intrinsic subtypes of tumors will facilitate better understanding of cancer etiology and more effective targeted cancer therapies. Transcriptional dysregulated programs, which dictate the expression of genes that define tumor phenotype and inherent heterogeneity are often driven by the function of crucial transcription factors. Identification and molecular characterization of such transcription factors that serve as major regulators of genes and pathways are likely to offer new insights into fundamental mechanisms of cancer biology as well as unveil novel avenues for therapeutic interventions in cancer. Here we have focused our studies on ETS1, an oncogenic transcription factor and examined its functional role in governing dysregulated transcriptional networks in specific subtypes of both Breast Cancer (BRCA) and Head and Neck Squamous Cell Carcinoma (HNSCC). We discovered that ETS1 is preferentially over-expressed in the Basal and Claudin-Low subtypes of BRCA as well as the Mesenchymal subtype of HNSCC, which are all aggressive forms of cancer that currently lack targeted therapy. By using a comprehensive molecular, genetic and bioinformatics driven analysis, we have identified direct transcriptional targets of ETS1 in cell-line models of the aforementioned cancer subtypes. Our integrative approach has uncovered core gene-signatures that highlight ETS1 as a crucial regulator of several oncogenic processes that are intrinsically linked to the phenotypes that define specific subtypes of BRCA and HNSCC. Collectively our studies have revealed the mechanisms by which ETS1 drives epithelial tumorigenesis and in a broader context, identified the key transcriptional networks to which ETS1high tumors are addicted for growth and survival. These findings can be exploited for rational cancer therapies by leveraging the knowledge of novel ETS1 targets that those tumors are likely to be vulnerable to, importantly in the context of a defined molecular tumor subtype.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Polyamine Biosynthetic Metabolic Dysregulation: Targeting and the Adaptive Response
No full textPh.D.Despite a 98.9% 5-year survival rate for prostate cancer (PCa), 20-30% of individuals will recur after the 5-year mark and will ultimately develop lethal hormone refractory disease, also known as Androgen Deprivation Therapy–Resistant prostate cancer (ADT-RPCa). Research in the field has concentrated on the idea that resistance is due to changes along the androgen axis, and therefore better targeting of this axis will reduce recurrence or prolong clinical benefit of ADT. We offer a novel approach for identifying and leveraging metabolic stress unique to PCa as a way to complement attacks on the androgen axis, while working via mechanisms independent of that axis. We question how to both exploit metabolic changes for therapeutic benefit and circumvent potential resistance mechanisms, and believe understanding adaptive response to metabolic therapy is key. We previously developed a bioinformatics pipeline to determine which metabolic pathways are most dysregulated based on analysis of transcriptomic data in PCa, and identified the polyamine biosynthetic pathway as being highly uniquely dysregulated in prostate cancer. Consequently, our lab has shown successful selective targeting of the polyamine biosynthetic pathway and connected methionine salvage pathway in PCa both in vivo and in vitro. Treating with the polyamine analogue, N(1),N(11)-bis(ethyl)norspermine (BENSpm), and inhibition of the methionine salvage pathway (MSP) with the methylthioadenosine phosphorylase (MTAP) inhibitor Methylthio-DADMe-Immucillin-A (MTDIA) eliminates a highly protected mechanism by which prostate cells relieve metabolic strain caused by increased polyamine biosynthesis, albeit to varying sensitivities. Therefore, the overall objective of this research is to identify the adaptive response that arises when polyamine catabolism is activated along with inhibition of methionine salvage in prostate cancer. Further, we hypothesized that identification of a Master Metabolic Transcriptional Regulator of the response to therapy represents potential mediators of adaptation that can be exploited to stratify PCa patient populations that will be most responsive to the drug combination being proposed, while also identifying a population that would benefit from new strategies targeting the adaptive response. Preliminary findings implicate the MMTR, JAZF1, as crucial to metabolic reprogramming in PCa. We propose to determine the role of JAZF1 in regulating metabolism of PCa and determine how this may impact response to therapeutic intervention. This novel approach takes advantage of an inherent metabolic strain accentuated in prostatic epithelial cells and enhanced in PCa, in order to develop novel therapeutic strategies and expands upon it by understanding how PCa cells attempt to adapt to metabolic therapy-induced stress.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Assessing the Global Occurrence, Fate, and Ecotoxicological Effects of Contaminants of Emerging Concern in the Environment Using Target Analysis and Suspect Screening
No full textPh.D.The occurrence of pharmaceuticals in the environment from discharges of wastewater treatment plants (WWTPs) and from the land application of antibiotic-laden manure from animal agriculture is a critical global issue because these residues have been associated with the increased emergence of antibiotic resistance in the environment. In addition, pharmaceutical residues may have adverse ecotoxicological effects in exposed aquatic organisms. This dissertation involves the optimization of different methods and workflows using liquid chromatography and mass spectrometry (LC-MS) to investigate the occurrence and fate of pharmaceutical residues in the environment, which are considered contaminants of emerging concern (CECs). Chapter 1 discusses an introduction on the sources of pharmaceuticals and contaminants of emerging concern (CECs) in the environment. The chapter also provides information on sample preparation and instrumentation techniques used for the analysis of CECs, as well as the scope of all the succeeding chapters in this dissertation. Chapter 2 presents the development of the ion ratios between quantifying and qualifying ions as a quality assurance criterion in the analysis of pharmaceutical residues in wastewater and surface water. Thirty pharmaceuticals were investigated and it was found that ion ratio variabilities increase at lower concentration and in matrices that are more complex. The variations of ion ratios at different concentrations should be considered because if only one tolerance limit is applied across all concentrations, it is highly likely that false negative results will occur at low concentrations, especially at around 1.56 to 12.5 ppb. The ion ratio tolerance range was optimized to be 40% for wastewater influent samples and 30% for wastewater effluent and surface water samples. A tolerance range of 80% was set for tetracyclines and quinolones that showed higher variations compared to other analytes studied. Chapter 3 shows the application of a robust targeted analytical method for the analysis of multiple classes of pharmaceuticals in six countries around the world. Results showed that pharmaceutical concentrations varied in the wastewater and surface water samples from Hong Kong, India, Philippines, Sweden, Switzerland, and the U.S. Ciprofloxacin and clarithromycin were the antimicrobials with the highest concentrations in all of the countries. Hong Kong showed the highest total antimicrobial concentrations that included macrolides, quinolones, and sulfonamides with concentrations reaching 60,000 μg/L levels in the influent. Antidepressants were frequently detected in Sweden, Switzerland, and the U.S. Chapter 4 investigates the removal efficiencies for antimicrobials, antidepressants, and other pharmaceuticals of seven WWTPs that employ varying treatment technologies. The concentrations of the pharmaceuticals were measured by LC-MS at multiple points during the course of treatment. Additionally, the ecotoxicological effects of the WWTP effluents were also evaluated based on their behavioral effects in exposed larval zebrafish. It was found that biological treatment process provided negative to low removal (95% overall removal). Notably, the final effluents of the seven WWTPs in this study did not show any significant behavioral alterations in zebrafish. In Chapter 5, suspect screening using LC with high resolution mass spectrometry (HR-MS) was used as an approach to expand the detection coverage of CECs in wastewater and surface water samples. Current workflows for suspect screening suffer from high frequency of false positives or lack of confidence in the true identity of the compounds detected. An optimized workflow was developed to determine the occurrence of CECs in wastewater and surface water samples from Hong Kong, India, Philippines, Sweden, Switzerland, and the U.S. Results revealed the presence of 61 contaminants that include pharmaceuticals, pesticides, and industrial chemicals. The workflow reported in this study provides an efficient approach and increased level of confidence in the identification of CECs using open access databases. In chapter 6, surface waters in both urban and rural sites in Bangladesh were collected and analyzed for the presence of antibiotic residues and other pharmaceuticals. Ciprofloxacin and clarithromycin were the antibiotics detected at the highest concentrations using a targeted LC-MS method. Using a suspect screening approach and retrospective analysis of the water samples with HR-MS, additional antibiotics clindamycin, lincomycin, linezolid, metronidazole, moxifloxacin, nalidixic acid, and sulfapyridine were detected. Prevalence of amoxicillin transformation products in surface waters was also confirmed. In addition, medicinal and agricultural antifungal compounds were frequently found in the samples.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Predictions for the Decay of the Higgs Boson to Bottom Quarks to Third Order in QCD
No full textPh.D.In this dissertation we present a fully-differential calculation of the decay of the Higgs boson to bottom quarks (H->bb) at next-to-next-to-next-to-leading-order (N3LO) accuracy in QCD. We regulate the infrared divergences present in the calculation at this perturbative order by using the Projection-to-Born technique coupled with the N-jettiness slicing method. In the first part of the dissertation we compute the 1-jettiness soft function at NNLO accuracy. This function is one of the required pieces in order to employ N-jettiness slicing to handle infrared divergences. The calculation uses the known expressions for the emission of one and two soft partons and is performed using a sector-decomposition approach that is spelled out in detail. Results are presented in the form of numerical fits to the 1-jettiness soft function for LHC kinematics (as a function of the angle between the incoming beams and the final-state jet) and for generic kinematics (as a function of three independent angles). In the second part of the dissertation we consider the NNLO corrections to the decay process H->bbj, where j represents additional QCD radiation. In this calculation we treat the bottom quarks as massless and focus on contributions in which the Higgs boson couples directly to bottom quarks through a non-zero Yukawa coupling. We compute the various components needed to construct the NNLO contribution, including an independent calculation of the two-loop amplitudes, and build a Monte Carlo implementation of H->bbj. Finally, in the third part of the dissertation we use the results of the H->bbj calculation to obtain fully-differential predictions for the H->bb decay at N3LO accuracy. We discuss how to combine the Projection-to-Born method with the N-jettiness slicing procedure and extensively validate our methodology. We present exclusive jet rates and differential distributions for jet observables at N3LO using the Durham jet algorithm in the Higgs boson rest frame.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Women Leaders in STEM Disciplines: A Comparative Case Study on Leadership and Disciplinary Culture
No full textPh.D.Women represent only 24.5% of positional leaders in academia (Lennon, 2013). The experiences of women in men-dominated domains vary by discipline (Becher & Trowler, 2001; Bystydzienski & Bird, 2006), and it is not known how women in STEM define leadership in their discipline and whether they understand these leadership expectations to be gendered. Understanding how women in STEM perceive leadership expectations can illuminate how women position themselves to attain and persist in leadership positions. This qualitative study used a single embedded case study to compare the experiences of women leaders in biology, a discipline that is dominated by women students, and reaching gender parity on the faculty level, with those in engineering, a discipline dominated by men students and faculty, with both groups of women leaders embedded within the larger context of academia. Individual interviews and document analysis were used to construct case studies. This study investigated the influence of a masculinist culture on the leadership of women in STEM in academia. The purpose of this study was to compare gendered leadership expectations between two disciplines: biology and engineering. Specifically, this study focused on the experiences of women leaders in mechanical engineering and computer science engineering, sub-disciplines with similar disciplinary cultures (Becher & Trowler, 2001; Gibbons, 2009; Yoder, 2016). In doing so, it addressed how gendered experiences in a masculinist culture shape women's leadership styles in different ways. The theory of gendered organizations was used to analyze and discuss results. This theory states that organizations are framed in the advantage and disadvantage of men and women, and that gender is reproduced and maintained in organizations through five processes (Acker, 1990). These processes include the maintenance of gendered divisions, the construction of symbols and images that represent those divisions, the presence of interactions between men and women that produce gendered social structures, the formation of individual identity through gendered processes and the creation of organizational logic, the underlying assumptions that are present in work organizations (Acker, 1990). Using this framework, three themes emerged from the case studies. Women in both disciplines experienced a masculinist culture, and adopted gendered leadership styles and gender-neutral narratives of leadership. However, more women in biology than engineering adopted the gendered leadership styles of empathetic leadership and collaborative and empowering leadership in a discipline with an increasing number of women and a disciplinary culture moving toward collaboration. The number of women in the discipline influenced how unencumbered women were to adopt gendered leadership styles and acknowledge and share gendered experiences. Women in biology were more able to use their gendered experiences in shaping their leadership style to empower and empathize with others. Women in engineering were required to demonstrate strength, toughness and a lack of emotion and felt more restricted by a masculinist culture, from adopting traits such as empathy in their leadership. Additionally, more women in engineering than biology adopted gender-neutral narratives of leadership, that were data driven, impersonal and the valued the scientist identity. Women in engineering sought to distance themselves from gender disparities in their leadership experiences in order to preserve gender-neutral narratives which emphasized a future in which both men and women needed the same skills and traits to succeed as leaders. These gender-neutral narratives conflicted with current realities of gendered interactions. In applying Acker’s theory of gendered organizations, the element of gendered organizational logic was evident as both women leaders who embraced gendered leadership traits and gender-neutral leadership traits faced limitations. Women who adopted gendered leadership that was empathetic, confident, empowering and collaborative faced the limitation of being perceived negatively by their peers, forcing them to manage the impressions of others so as not to be perceived as weak, indecisive or overly confident. Women who adopted gender-neutral leadership traits to maintain solidarity with their men peers, faced the limitation of being expected to be better communicators than men and to not surpass men in strength or decisiveness. Whether women adopted gendered leadership or gender-neutral leadership they faced gendered leadership expectations and modified their behavior and identity to navigate those expectations.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*