50107 research outputs found
Sort by
On Faces of Flow Polytopes
No full textThis thesis will study the -vectors and facial structure of two families of flow polytopes: those coming from complete graphs and those coming from grid graphs.
In the first part of this thesis, we study flow polytopes coming from the family of complete graphs , which include the famous Chan-Robbins-Yuen polytope (). We give generating functions and explicit formulas for computing the -vector by using Hille's (2003) result bijecting faces of a flow polytope to certain graphs, as well as Andresen-Kjeldsen's (1976) result that enumerates certain subgraphs of the directed complete graph. We extend our results to flow polytopes of the complete graph having arbitrary (non-negative) netflow vectors and recover the -vector of the Tesler polytope of Mészáros--Morales--Rhoades (2017).
In the second part of this thesis we study generalized Pitman-Stanley polytopes (), which we show are flow polytopes of a grid graph for certain netflow vectors. We study the vertices of this family of polytopes, and provide descriptions of such in terms of integer flows. Moreover, we give recursive formulas for the number of faces, as well as generating functions for the number of vertices, and prove that---for a fixed and ---these generating functions are a polynomial in .
In the third part of this thesis, we investigate the face lattice of both of these families of polytopes, giving a recursive description of the face lattice of generalized Pitman-Stanley polytopes when . Moreover, we investigate a family of polytopes closely related to flow polytopes---fractional perfect matching polytopes---and find a formula for the number of vertices of the fractional perfect matching polytope of the complete graph.Doctor of Philosophy (Ph.D.)2026-05-1
Engineered Salmonella for Antigen-Independent Cancer Immunotherapies
No full textImmunotherapies have revolutionized cancer treatment; however, they do not work for all patients. Today’s immunotherapies, including checkpoint blockade and CAR-T cells, rely on knowing a patient’s specific tumor mutations to target treatments. Developing off-the-shelf tumor agnostic immunotherapies would overcome barriers to effective treatment including cost and development time. Salmonella locate to the tumor microenvironment with high specificity, are innately immunostimulatory, and can produce protein products – making them a promising treatment vehicle.
In this dissertation, I engineered Salmonella using three approaches to develop tumor agnostic therapies. First, immune controlling Salmonella were engineered to controllably produce surface lipopolysaccharides through tuned expressed of the gene msbB to increase their immunogenicity. Production levels were controlled by distinct ribosomal binding sites and protein degradation tags to produce an effective dose below the toxicity threshold. Controlled lipopolysaccharide production activated monocytes in vitro and in vivo increasing TNFα production without inducing mouse illness. Immune controlling Salmonella reduced tumor growth and increased survival. Second, Salmonella were engineered to intracellularly deliver the COVID-19 spike protein to cancer cells to redirect vaccine immunity. In this work, I engineered a fusion protein of the most immunogenic regions of the spike protein. Delivery of this fusion protein activated spike-specific T cells from prior vaccination. In vivo, spike protein delivery by Salmonella reduced tumor growth and activated antigen presenting cells to generate further anti-tumor immunity. I developed a mathematical model of tumor immunotherapy using antigen delivering Salmonella to elucidate the interactions between Salmonella, cancer cells, and T cells. In this model, I observed that a balance between intracellular and extracellular Salmonella is critical for effective therapies. Thirdly, I developed a technique of bacterial pre-targeted radioimmunotherapy for the localization of the radiohapten DOTA to tumors. I engineered Salmonella to express DOTA binding antibody fragments on their surface. In this project, I tested three systems for surface display and determined that using an OmpA fusion protein strongly bound to DOTA in vitro with similar kinetics to the free antibody.
Each of these antigen agnostic therapies targets treatments to tumors without knowledge of the tumor mutations, overcoming a major limitation of traditional immunotherapies.NSF GRFP: 1938059
DoD: HT94252310067
NIH: R21CA287211Doctor of Philosophy (Ph.D.)2026-05-1
FRACTURED LOCALISM: HOW DIVIDED TIMES MAKE CONTENTIOUS PLACES
No full textResponding to increased national political and cultural division, a flood of scholarship has addressed the roots of political polarization, the development of Americans’ political beliefs and animosity towards one another, and solutions to mitigating extreme division. Local and national journalists have reported on large-scale trends of political polarization and its threats to our institutions, as well as communities that are divided, not just over candidates, but also community values. While social and symbolic boundary-making, particularly across race and class, has long been the subject of sociological work in changing cities and neighborhoods, few have examined how political division structures community life. Through ethnographic observation, in-depth interviews (n=68), and photographic data in a politically heterogeneous gentrifying small city, I examine how national division manifests in everyday life, addressing the question of how community is achieved in divided times. After identifying four main subgroups of Beacon, NY’s social landscape, the first chapter examines how local conflict serves as a fulcrum for boundary-making along lines of race, class, and politics, and the strategies community members use to establish a sense of belonging. The second chapter focuses on local placemaking, proposing a scriptive urbanism framework to show how oldtimers and newcomers, conservatives and progressives navigate the city’s divisive ‘place identity crisis:’ through inscription (i.e., shaping the city’s physical form), ascription (i.e., imputing meanings on place), and subscription (i.e., adopting or rejecting place character). The third chapter moves from division to places of solidarity that facilitate cross-cultural connections thereby mitigating local fractures across differences, proposing that these ‘recognition zones’ serve as vital community resources in these uncertain times. Ultimately, this dissertation demonstrates the importance of examining urban political culture as a facet of community change, understanding the local as a stabilizing anchor, and developing strategies for mending local fractures through recognition in divided times.UMass Amherst Graduate School
UMass Amherst Sociology Department
American Sociological Association's Community and Urban Sociology SectionDoctor of Philosophy (Ph.D.)2026-05-1
INVESTINGATING THE ROLE OF SENSORY INFORMATION ON NEURAL ACTIVITY IN THE SOCIAL BEHAVIOR NETWORK
No full textCoordination between members of the same species is essential for survival and underlies behaviors such as reproduction, protection from predation, food allocation and offspring care. Such coordination requires that individuals accurately identify social partners or social cues and respond efficiently. These tasks are orchestrated by a network of brain regions called the social behavior network (SBN). This network generates social behavior from socially relevant sensory input, but how different populations of neurons in the widely conserved network orchestrate this on a moment-to-moment basis is still largely unknown. In this dissertation I present four chapters and appendices that seek to uncover how the brain orchestrates socially relevant sensory processing throughout the SBN. Chapter one synthesizes prior research and scientific advances to propose novel approaches to better understand integrative social behavior circuits. Chapter two explores the impact of the social environment on neural activity of accessory olfactory bulb (AOB) neurons. Predator stimuli increase response to social stimuli in both sexes and direct contact induces oscillatory activity. Chapter three extends these investigations to an output of the AOB: aromatase-expressing neurons in the medial amygdala (MeA).
Results show that these neurons respond robustly to various social stimuli (adult males, females, and pups) but not directly to predators. Instead, predator presence indirectly enhances responses to female stimuli in males. Additionally, these neurons display strong novelty responses with rapid adaptation, with females exhibiting stronger adaptation to same-sex stimuli. Advances in computational ethology (pose-estimation and behavioral modeling) allowed us to analyze moment-to-moment social interactions and reveal that MeA aromatase neurons are most active in male-male conditions, with age and isolation influencing female neural activity. The same methods within female bed nucleus of the stria termanilis estrogen receptor alpha (BNST ER alpha) neurons, displayed a preferential response to male stimuli – a pattern attenuated in receptive hormonal states (Fig. 33-36). Appendix A synthesizes prior research to better understand the role of aromatase in the brain in establishing and maintaining sex differences in neural circuits. Collectively this work elucidates neural processing patterns within different SBN nodes, illuminating functional roles and their ability to adapt in real-time.Doctor of Philosophy (Ph.D.)2026-05-1
Accelerating River Discharge in High Mountain Asia
No full textHigh Mountain Asia (HMA) plays a crucial role in Asian hydrology—its vast snow and glaciercovered landscape significantly influences downstream river water supply for billions of people. Understanding the spatiotemporal pattern of river discharge in HMA aids effective water resource management and infrastructure planning. In this study, we used a combination of hydrologic modeling and assimilation of remotely sensed discharge from Landsat and PlanetScope imagery to investigate how daily river discharge has changed for more than 114,000 reaches across HMA between 2004 and 2019. We observed significant increasing trends in river discharge for 11,113 reaches (∼10%), particularly in smaller rivers of the Syr Darya, Indus, Yangtze, and Yellow River basins. The ratio of total glacial melt and precipitation received by individual river reach showed an average significant increase of 2.2% per year, particularly in the Syr Darya, Amu Darya and Western Indus rivers. Across HMA, our results also indicate that 8% of river reaches with either planned and existing hydropower plants or dams experienced a statistically significant average increase of 2.9% per year in stream power. These findings illustrate the rapidly changing patterns of river discharge and stream power in HMA
AN APPLICATION OF MDO TO LOW INDUCTION ROTORS USING HIGH FIDELITY STRUCTURAL MODELING TOOLS
No full textUpscaling wind turbines has been an important technological trend in the past decade. For rotor design, this has led to the incorporation of structural and cost analyses to realize feasible designs, contrary to previous approaches in which the main design driver was aerodynamic performance. This evolution of the rotor design problem has made it a multi-disciplinary design optimization (MDO) problem, receiving attention from the scientific community in the form of new, MDO design frameworks and tools. As the upscaling trend develops, new design objectives have been proposed to this design problem. As a result, alternative approaches to the Betz optimal-oriented design have been introduced, such as the Low Induction Rotor (LIR). The LIR, which includes blade radius within its optimization, introduces a structural constraint in the form of root blade bending moment M0. This intrinsically multidisciplinary problem has not been sufficiently tested in MDO environments, usuallyl acking mid to high fidelity analysis in the structural discipline. This master thesis reviews the LIR design theory and demonstrates its application within the context of MDO, also presenting an alternative constraint in the form of maximum tip deflection δt. The optimization problem is formulated by leveraging the use of a parameterized axial induction distribution along the blade span, which provides a set of free variables. The set of free variables is changed in a range to materialize variations of traditional LIR theory, allowing a design space exploration to obtain pareto optimal families of blades, alternate to the IEA 15 MW reference turbine.Wind Energy CenterMaster of Science (M.S.
SYNTHESIS OF HIERARCHICAL BEA ZEOLITE BY POST SYNTHETIC SURFACTANT TEMPLATING PROCESS
No full textZeolites are porous aluminosilicates used as catalysts in several industrial reactions. They have wide applications due to their crystalline structures with intrinsic microporosity (< 2 nm). These intrinsic small porous structure leads to slower transport of bulky species which causes diffusion limitation, unwanted side reactions and deactivation of catalysts. Overall efficiency of the catalysts is therefore reduced. To solve the slow mass transport in zeolite catalysts we incorporate mesopores (2-50 nm) in the framework of zeolites thus forming hierarchical zeolites which are proven as an effective method for addressing the transport issue. Currently two techniques are used for incorporating mesopores –bottom up and top down. In bottom-up method mesopores are created along with micropores during their synthesis, but it is economically impractical because of costly templates and altered synthesis routes. In top-down methods, specifically surfactant templating method is efficient, easy and flexible with respect to incorporation of mesopores, as we can largely vary the process parameters for a wide range of zeolites and control the mesoporosity.
This research focuses on surfactant templating of *BEA (Beta) zeolite. The aim is to study how surfactant templating methods differ in aluminum containing *BEA (Al-BEA) and de-aluminated *BEA (D-BEA). D-BEA was synthesized from dealumination of commercial Al-BEA in acidic condition. Both D-BEA and Al-BEA were surfactant templated using a wide range of sodium hydroxide (NaOH) and surfactant concentrations. XRD results assured that the crystal structure of the zeolites can be maintained after the surfactant templating process. N2 adsorption/desorption isotherms were used to characterize textural property of the hierarchical zeolites. It was observed that the BET surface area of D-BEA increased up to 800 m2/g from 500 m2/g with an optimized surfactant templating treatment. NLDFT pore size distribution analysis also confirmed the incorporation of mesoporosity in both Al-BEA and D-BEA samples. It was found that the parent Al-BEA had an external surface area of around 186.9 m2/g which was increased to 438.9 m2/g. Parent D-BEA had an external surface area of around 145.3 m2/g which was increased to 808.4 m2/g after surfactant templating.Master of Science in Chemical Engineering (MSChE
Peer Interaction in Chinese as a Second Language: The Interplay of NoM, LREs, and Translanguaging
No full textWhile much research has been conducted on interaction in second language acquisition, the interplay of key interactional mechanisms—Negotiation of Meaning (NoM), Language-Related Episodes (LREs), and translanguaging—remains underexplored in the context of Chinese as a Second Language (CSL). With Mandarin gaining prominence globally, understanding how CSL learners use these strategies during peer interaction is crucial for enhancing classroom practice and informing language pedagogy. This study investigates the development and co-occurrence of NoM, LREs, and translanguaging in the peer talk of eleven CSL learners enrolled in three proficiency-based university classes (novice-low, novice-high, and intermediate-low).
The study uses audio recordings of collaborative tasks—information-gap, decision-making, and cultural discussions—conducted in natural classroom settings. Peer interactions were transcribed and analyzed through both qualitative discourse analysis and quantitative comparison. The analysis focused on the frequency and types of interactional strategies, patterns of co-occurrence between NoM and LREs, and the functional and affective uses of translanguaging. Key measures included learners’ use of clarification requests, self-corrections, lexical or grammatical focus, and the integration of first-language resources into discourse.
Findings reveal a developmental trajectory in which learners progress from relying heavily on lexical negotiation and basic translation to demonstrating greater metalinguistic reflection and emotional regulation through translanguaging. At the intermediate level, over 75% of NoM instances were immediately followed by LREs, showing how learners used moments of communicative uncertainty to collaboratively reflect on language form. Additionally, translanguaging evolved from a tool for bridging gaps to a resource for managing discourse and expressing stance. These changes suggest a growing capacity for autonomous, socially situated language use.
In contrast to prior studies that rely on scripted or experimental tasks, this research is grounded in spontaneous peer interaction, capturing the authentic dynamics of bilingual strategy use in real-time learning. The findings emphasize the importance of recognizing the interconnectedness of interactional strategies and creating classroom environments that support exploratory, collaborative talk. Ultimately, this study highlights how learners’ strategic integration of NoM, LREs, and translanguaging contributes to both linguistic development and classroom participation, offering practical implications for CSL educators seeking to foster meaningful peer interaction.Master of Arts (M.A.
Advance Care Planning Conversations in Primary Care
No full textBackground: Advance Care Planning (ACP) conversations that occur late in the course of illness often have detrimental effects on the quality of life, and cause distress to patients, families, and healthcare providers. Research supports the need for early initiation of ACP conversations in noncrisis situations. A key barrier is the lack of healthcare providers’ knowledge, comfort level, and confidence in initiating such conversations. The purpose of this DNP project was to increase the beliefs, knowledge, and self-efficacy of Nurse Practitioners (NP) in initiating ACP conversations during routine care appointments.
Methods: A pretest-posttest design was implemented. Bandura’s social cognition theory provided a framework for intervention development focusing on provider education to create change. Nurse Practitioners were recruited to participate through the Massachusetts Coalition of Nurse Practitioners (MCNP) listserv. Those who agreed to participate were provided an online PowerPoint presentation that focused on having ACP conversations. They were asked to complete both pre and post intervention surveys to measure their current beliefs, self-perceived knowledge, and self-efficacy regarding ACP discussions. The quantitative data obtained from the surveys was analyzed using descriptive statistics.
Results: The mean score of beliefs, self-perceived knowledge, and self-efficacy regarding ACP discussions were compared pre and post intervention with noted improvement in all responses post intervention except for the question “Patients receive care at the end of life that is congruent with their wishes” which remained the same with most responses being “somewhat.”
Conclusion: This project supports that an evidence-based educational intervention can have a positive impact on provider’s beliefs, knowledge, and self-efficacy and improve provider’s ability to perform ACP conversations during routine appointments in the future.Doctor of Nursing Practice, Family Nurse Practione