DSpace@RPI (Rensselaer Polytechnic Institute)
Not a member yet
6809 research outputs found
Sort by
Enabling Cross-Language Data Integration and Scalable Analytics in Decentralized Finance
No full textWith the agile development process of most academic and corporate entities, designing a robust computational back-end system that can support their ever-changing data needs is a constantly evolving challenge. We propose the implementation of a data and language-agnostic system design that handles different data schemes and sources while subsequently providing researchers and developers a way to connect to it that is supported by a vast majority of programming languages. To validate the efficacy of a system with this proposed architecture, we integrate various data sources throughout the decentralized finance (DeFi) space, specifically from DeFi lending protocols, retrieving tens of millions of data points to perform analytics through this system. We then access and process the retrieved data through several different programming languages (R-Lang, Python, and Java). Finally, we analyze the performance of the proposed architecture in relation to other high-performance systems and explore how this system performs under a high computational load
Advances in the theory, implementation, and application of mechanistic models in downstream bioprocessing of monoclonal antibodies
No full textJuly 2024School of EngineeringMechanistic models are powerful tools for process characterization and optimization. Although their usage in academia is common, their widespread use in the biopharmaceutical industry is hindered by several factors. First, derivation of first-principles models is very difficult, sometimes due to insufficient understanding of the underlying phenomena or limitations in the experimental techniques or instruments. Deriving empirical models also requires instinct and experience in relevant fields. Second, there are many unit operations involved in the production process. There is a lack of a powerful, yet flexible and shared platform to perform these simulations. Third, the workflow for using these developed models is not clear. Applying these models in an already established industry workflow and interpreting the results require advanced knowledge in multiple disciplines. In this thesis, we aim to address these challenges, particularly for protein A chromatography and precipitation capture purification operations in the downstream bioprocessing of monoclonal antibodies. In protein A chromatography, we derived a novel isotherm model and implemented it with the general rate model in well-known chromatography modelling software packages including GoSilico and CADET. Experiments are designed and performed to demonstrate that the model is compatible with an array of resins, mAbs and experimental conditions. We show that the model can be easily trained and that it offers excellent chromatogram predictions. We further established an intuitive workflow showing how to use the model for process development. Beyond these practical applications, we also used models of varying complexity to understand the heterogeneous surface properties of the protein A resins and tried to determine what compromises are needed when applying complicated semi-empirical and first-principles mechanistic models to practical situations where speed, simplicity and accuracy are required simultaneously. In precipitation, we focused on the solution, implementation and benchmarking of a mechanistic model for solid particle formation kinetics called the population balance model in the free and open-source software package CADET. CADET already supports the solution of many models for other unit operations, and introducing population balance model to the CADET family drives integrated in silico process development. In addition to this, we also applied the model to characterize antibody precipitation kinetics: experiments were conducted and a workflow was established to apply the model. The results were further used to inform future process development directions.Ph
Time-resolved optical tomography platform for mesoscopic lifetime imaging
No full textMay 2024School of EngineeringContinuous progress in understanding oncogenesis has driven the development of targeted therapies, which leverage the targeted drugs binding to their specific receptors. However, the clinical success of these therapies has faced challenges such as drug resistance, which has been found to be associated with spatial variation in tumor regions, as well as insufficient drug exposure to the affected cells. During drug development, preclinical in vivo molecular imaging is a crucial tool for facilitating the drug discovery pipeline by assessing the drug's biological biodistribution and pharmacodynamics in live intact subjects. Molecular imaging modalities such as positron emission tomography (PET) and optical imaging provide the means to monitor biological processes at the molecular level during targeted treatment. Mesoscopic imaging is a technique used to examine the intricate details of biological systems within an organ or body part. This approach is particularly useful in preclinical cancer research as it allows for the investigation of the interaction between tumors and their surrounding tissue environment, which is much more complex than just a simple collection of cells. However, it is still challenging to noninvasively monitor the intra-tumoral heterogeneity (spatial variation in tumor regions) and assess the drug target engagement (the degree of drugs binding to their specific receptors) during drug delivery at the mesoscopic scale (a few hundred micrometers of resolution and few millimeters imaging depth). Indeed, optical imaging is unique in allowing the monitoring of multiple biomarkers simultaneously while providing structural, functional, and molecular contrasts through various light-tissue interactions. Forster Resonance Energy Transfer (FRET) imaging technique can sense protein-protein interaction processes, such as the binding of targeted antibodies to their respective receptors, through approaches such as the Fluorescence Lifetime Imaging (FLI) modality. On the other hand, fluorescence molecular tomography (FMT) has demonstrated its utility in visualizing the three-dimensional (3D) distribution of fluorescent molecular probes. Considering the binding activity of antibodies and the targeted receptors, intra-tumoral heterogeneity can be reported by retrieving the distribution of the fluorescent-labeled antibodies. This project aims to develop a noninvasive optical imaging platform capable of monitoring target engagement and intratumoral heterogeneity in the mesoscopic regime. With the integration of two imaging modalities -- fluorescence lifetime imaging via Forster Resonance Energy Transfer (FLI-FRET) and fluorescence molecular tomography (FMT) at the mesoscopic scale, we developed a custom-built optical imaging system, Time-domain Mesoscopic Fluorescence Molecular Tomography (TD-MFMT), which enables assessing target engagement via FLI-FRET as well as quantifying 3D distribution of multiplexed fluorescent-labeled antibodies through FMT modality. Hence, the outcomes of this project represent a significant advancement in optical molecular imaging for preclinical research and development.Ph
Eliciting Survey Knowledge with Semantic Data Dictionaries
No full textMany countries perform surveys to gather data from their population for supporting decision-making and development of public policies. Questionnaires are possibly the most used type of data acquisition instrument in surveys, although additional kinds may be employed (especially in health-related surveys). In the United States, the NHANES is a national health and nutrition examination survey conducted by the National Center for Health Statistics, designed to collect data on adults' and children's health and nutritional status. Data is organized in several tables, each containing variables to a specific theme, such as demographics, and dietary information. In addition, data dictionaries are available to (sometimes partially) document the tables' contents. While data is mostly provided by survey participants, instruments might be collecting data related to other entities (e.g. from participants' households and families, as well as laboratory results from participants' provided blood and urine samples). All this complex knowledge can often only be elicited by humans when analyzing and understanding the data dictionaries in combination with the data. The representation of this knowledge in a machine-interpretable format could facilitate further use of the data. We detail how Semantic Data Dictionaries (SDDs) have been used to elicit knowledge about surveys, using the publicly available NHANES data and data dictionaries. In SDDs, we formalize the semantics of variables, including entities, attributes, and more, using terminology from relevant ontologies, and demonstrate how they are used in an automated process to generate a rich knowledge graph that enables downstream tasks in support of survey data analysis
Design of advanced ceramics for extreme corrosion environments
No full textMay 2024School of EngineeringCorrosion in the case of ceramics is the destructive attack of a ceramic by chemicals within its environment, which, based on the application of the ceramic, may lead to very expensive repair and replacement costs to maintain functionality. Our work here focuses on investigating the underlying corrosion mechanisms of (1) ceramic nuclear waste form materials used for the immobilization of long-lived actinides and (2) Environmental Barrier Coatings (EBCs) used to provide jet engine components with protection from harsh, high temperature, combustion environments. Firstly, accelerated leaching tests were used to measure the chemical durability and identify the leaching mechanisms of amorphous, alpha, beta, and gamma phase yttrium disilicate. Rare earth disilicates may be formed in geological repositories from the reaction of waste actinides and smectite found in clay barriers. Hence, understanding their corrosion mechanism could aid in predicting their long-term retention of actinides ions. Next, high entropy/multicomponent titanate pyrochlores and zirconate fluorites were synthesized as potential host materials for nuclear waste. Accelerated leaching tests were used to measure their elemental release rates and investigate their leaching mechanism. Our results showed that radius size disorder significantly impacts the chemical durability of multicomponent oxide solid solutions. By implementing more controlled design parameters, improvements in the corrosion resistance of these materials may be achieved. The latter part of our work focuses on the design and testing of multicomponent rare earth phosphates as potential EBC candidates. We observed superior molten glass corrosion resistance of (Lu0.2Yb0.2Er0.2Y0.2Gd0.2)PO4 and (Lu0.2Yb0.2Er0.2Dy0.2Gd0.2)PO4 at 1300 °C compared to single component phosphates, with Scandium appearing to have a negative impact of the corrosion resistance of the multicomponent phosphates. Similarly, at 1400 °C (Lu0.2Yb0.2Er0.2Y0.2Gd0.2)PO4 and (Lu0.2Yb0.2Er0.2Dy0.2Gd0.2)PO4 samples showed the least recession of their bulk material after molten glass corrosion. Our water vapor corrosion experiments at 1400 °C revealed the improved corrosion resistance of our designed multicomponent rare earth phosphates compared to LuPO4. Where the multicomponent phosphates showed less mass loss and less porosity increase after 60 hours compared to LuPO4.Ph
Understanding dynamics of virus propagation in plate and suspension cell cultures to improve upstream biomanufacturing performance
No full textAugust 2024School of EngineeringVirus replication is pivotal in two broad aspects of biomanufacturing: detecting viral adventitious agents and producing therapeutics such as vaccines and viral vectors. Although rare, viral contamination poses significant risks to patient safety and has substantial economic consequences, necessitating routine virus testing at various production stages. The current state-of-the-art for general virus screening is the in vitro virus (IVV) assay, which is labor- and time-intensive. In the first aim, a discrete model coupled with a partial differential equation solver was employed to simulate a monolayer of CHO-K1 cells and reovirus-3 diffusion in an IVV system. Model predictions and analysis indicated that virus transport to the cell surface is a major limitation, resulting in a lower initial infection rate than previous works have modeled. Model output was used to enhance the sensitivity of laser force cytology (LFC), an alternative detection technique. Results suggested that the current sampling rate may be too low to distinguish infection accurately. The second aim focuses on biomarker discovery for adventitious virus detection by examining miRNA and mRNA expression in infected CHO-K1 cells. Nine miRNA biomarkers of infection were identified across a panel of viruses (Reo-3, MVM, EMCV, and PIV-2) for general detection. Gene prediction analysis of differentially expressed miRNAs in early infection suggested their involvement in pro-viral mechanisms, while miRNAs in late infection were implicated in cellular stress responses. Moreover, miR-484, miR-409-3p, miR-450b-5p, and miR-21-3p were differentially expressed as early as day 1 post-infection in some viruses. This is well ahead of the appearance of virus-induced morphological changes in cells that are typically used for detecting infection. The third aim addresses optimizing virus yields in suspension cell culture-based vaccine production via mathematical modeling. A model of recombinant measles virus production in Vero cells predicted the optimal time of harvest for five bioreactor runs despite significant run-to-run variation. Sensitivity analysis indicated that virus attachment parameters and the degradation rate of the virus influence infection dynamics across the bioreactor run. Moreover, the model underscores the importance of accurately characterizing seed virus quality, particularly for defective interfering particle (DIP) content, to reduce run-to-run variation in production and maximize yield. Overall, this work advances the understanding of viral kinetics, providing critical insights for enhancing adventitious virus detection methods and optimizing virus production processes in biomanufacturing.Ph
The interstices of pain and relief: people, technology and governance in a palliative care social movement in india
No full textAugust 2024School of Humanities, Arts, and Social SciencesKerala, a state in India, has garnered acclaim as one of the more successful palliative care social movements by highlighting its functioning through a mix of community and state action. Kerala’s palliative care movement is unique in that it is a grassroots initiative that provides palliative care, free of cost, to people through a network of volunteers, local self-governing institutions and civil society at large This dissertation investigates the palliative care movement, its formation, and constituent elements, through the lens of social movements as studied within the field of Science and Technology Studies. This is done by focusing on three major aspects of the movement: community, governance, and technology. ‘Community’ here refers to the local yet contested conceptions of the term as it exists within the palliative care movement and across different regions of the state in Kerala. Governance refers to the different national and transnational antecedents that have shaped healthcare in Kerala and in turn has shaped the palliative care social movement. Finally, the painkiller morphine, widely understood as the primary instrument of pain relief shall be studied through its position as the most salient technology of care within the movement. This refers to the different political configurations that its use engenders, constrained as it is by transnational and national laws of regulation. Through a re-articulation of the theory of affordances, I show how each of the aspects of community, governance, and technology brings about unique affordances that allowed the social movement to take shape and sustain itself. I define affordances as actions that enable or constrain, made available through a combination of the socio-material realities of place, ethical positions, institutional histories and people in their multiplicity. Volunteering within palliative care organizations allowed volunteers to reassess and reshape how medicine and care can be imagined and disbursed. I further delineate how the volunteers’ own ideas of care were shaped and transformed by their own histories of organizing, caregiving, religious morals all together to form complex heterogeneous sources of inspiration and knowledge that they fed back into palliative care. This dissertation also highlights how the State of Kerala instituted new collaborative modes of governance that developed from state-led decentralization in the mid-1990s, which produced new affordances that palliative care organizations, both public and private, have capitalized on. Finally, through archival research, I contextualize the history of morphine regulation in India which constrains the delivery of pain relief medication to this very day. Further this dissertation also highlights how allowing volunteers to disburse morphine under a licensed doctor frees up some of these regulations in ways that people who need morphine do get it without the risk of misuse. Here again, affordances of this practice reimagine palliative care in a fundamental way. Ultimately, through a mix of ethnographic and historical research, I show how a Science and Technology Studies analysis of social movements can be carried out by considering different vantage points through the heuristic of the multiple affordances that enable and constrain them.Ph
Essays on ownership structure and corporate governance
No full textAugust 2024School of ManagementThis dissertation consists of two distinct but related essays examining the impact of ownership structure on corporate governance.The first essay investigates the relationship between passive institutional ownership and the firm’s pay disparity. Public concern regarding the gap between top managers and rank-and-file workers is rapidly growing with the Pay Ratio Disclosure rule’s enaction. This essay uses the direct pay ratio information disclosed by the firm’s proxy statements and the 2SLS approach to examine whether ownership structure impacts a firm’s pay disparity. More specifically, I investigate whether and how passive institutional investors affect the within-firm pay gap. I find that passive ownership is positively associated with the firm’s pay disparity. I also find that passive investors are more likely to support management regarding say-on-pay voting. Moreover, I also find that the increasing effect is more pronounced for firms with more talented CEOs and fewer active investors, and is positively associated with firm performance. The findings suggest that passive investors value managers’ ability more than pay disparity and influence a firm’s compensation policy through “behind-the-scene” engagement.
In the second essay, I examine the impact of passive institutional investors on corporation managerial behavior. Specifically, I exploit the Russell Index reconstitution to examine how passive investors impact firms’ forward-looking statements. I find that firms with more passive ownership modify their forward-looking statements within the MD&A year over year less. The findings indicate that an exogenous increase in passive ownership results in management releasing less new information in their forward-looking statements within the MD&A. Further analysis shows that the declined informativeness of the firm’s forward-looking statements is associated with higher stock crash risk and less firm-specific information in the firm’s stock price. The findings indicate that passive investors have a negative impact on the informativeness of the firm’s forward-looking statements within the MD&A and may preempt the firm’s forward-looking information from other channels of corporate disclosure.Ph
Engineering of human papillomavirus l1 virus-like particle to use as a scaffold for vaccines
No full textMay 2024School of ScienceVaccines are powerful therapeutics used to prevent and eradicate deadly infectious diseases, such as smallpox, and to reduce diseases such as polio, measles, tetanus, and more recently, the SARS-CoV-2 virus that caused the COVID-19 pandemic. Vaccines provide active acquired immunity toward a specific infectious disease, where the vaccine closely resembles the antigen, so that the immune system can recognize it as foreign, attacks it, and remembers it. This "remembering" is what makes vaccines so powerful; anything that resembles the antigen will be destroyed by the immune system in the future, providing long-lasting immunity. The first few vaccines developed were inactivated or attenuated viruses. Most of the vaccines in the world are usually in the form of subunit protein and adenovirus vaccines, but more recently, because of the COVID-19 pandemic, the emergency usage of mRNA-based vaccines has been implemented.Vaccine technology can also be applied in a nontraditional fashion, such as a contraceptive vaccine, where the vaccine target is pregnancy instead of a disease, to address the unmet need of a non-hormonal and non-invasive form of contraception. Therefore, we propose to design and engineer the Human Papillomavirus (HPV) L1 Virus-Like Particle (VLP) to be used as a scaffold for vaccines by inserting desired antigens. This means that this scaffold can be used for essentially any pathogen. Any peptide-based antigens can be engineered onto the surface of the VLP so that a vaccine against that antigen can be made. We propose to use VLP as a scaffold for a unique contraceptive vaccine, where the purpose of the vaccine is to prevent pregnancy. We also propose to use the VLP as a scaffold for a glycan-free COVID vaccine to minimize the occurrence of autoimmune diseases. Through this process, we have discovered ways to optimally assemble the VLP in our expression and production system. Lastly, we have engineered a small VLP, which is half the size of regular HPV VLP, that can also be utilized as a scaffold for different purposes with the intent of providing shorter immunity in the context of a contraceptive vaccine so that people will have different options for contraceptive duration.
This work focuses on the design of the VLP scaffold to accommodate different epitopes for a unique contraceptive vaccine and a glycan-free COVID vaccine, the methods to optimally produce the vaccines, and the ways to engineer a small-sized VLP to also be used as a scaffold.Ph
Flow physics and control of separation for low aspect ratio swept and tapered wings
No full textDecember 2023Flow separation is the last thing you want to happen over an aircraft wing: in most cases attached flow over the wing is the only thing keeping the plane in the air. The formation of 3D separated flows over wings and other control surfaces is an under-explored topic due to difficulties measuring or computing these phenomena. Despite this dearth in the literature, aircraft wings routinely encounter high-angle-of-attack scenarios either from transient gusts or intentional maneuvers. Unmanned aerial vehicles tend to operate at high angles of attack and low Reynolds numbers. The present experimental investigation seeks to provide a basis with which 3D separation can be characterized and controlled. Wind tunnel and water tunnel experiments were performed to show the effects of planform shape on the large scale flow structures over finite wings at high angles of attack. First, geometrically simple unswept planforms were considered. Wind tunnel experiments explored unswept planforms at the moderate Reynolds number , showing that these wings experience the owl-eye stall cell pattern at high angles of attack. The stall angle of attack increases with decreasing aspect ratio, as the end effects of the tip vortex and the horseshoe vortex at the wall get moved closer together and encourage attached flow. The surface foci of the stall cell are connected by an arch vortex which grows into the wake according to the spanwise distribution of the reversed flow region. At the midspan, the wake has some periodicity through the quasi-2D shedding of spanwise vortices into the wake. Next, the geometric complexity was systematically ramped up with a series of wind tunnel experiments on swept and tapered planforms at the moderate Reynolds number . For the untapered unswept wings, the stall cell pattern forms; for the swept back untapered wings the root-to-tip ram's horn forms, for the wings with forward swept trailing edge the inverted (tip-to-root) ram's horn forms and for swept back and tapered wings the ram's horn type surface spiral would form, though the 3D flow field shows some similarity to the arch type separation. When the leading edge is swept back, the reversed flow region and peak unsteadiness is shifted towards the wing tip, while when the trailing edge is swept forward, the reversed flow region and peak unsteadiness is shifted towards the root. Nine model wings were explored in water tunnel experiments at the very low Reynolds number . Some analogous topological patterns were identified between the wind tunnel and water tunnel tests in time-averaged volumetric velocity fields. Namely, the area of the reverse flow region as a function of the span showed very similar distributions at and as a function of sweep angle and taper ratio. The global flow structures like reverse flow region and spiraling 3D streamlines identified a ram's horn vortex for the swept back wing at both Reynolds numbers and an inverted (tip-to-root) ram's horn for the forward swept trailing edge wing at both Reynolds numbers. Finally, flow control using leading edge steady blowing with seven equally spaced rectangular jets was applied to two of the tapered wings in the wind tunnel at . Both wings had the same taper ratio and aspect ratio, but one had a forward swept trailing edge and the other had a swept back leading edge. The flow control for the trailing edge swept forward model was very effective, with beneficial aerodynamic augmentation being achieved when blowing is applied near the midspan. The flow control for the leading edge swept back model was less effective, as the leading-edge-normal blowing for the swept back leading edge angled the jet in a way that would increase the drag and fail to reattach the flow.Ph