Browsing by Subject "model"

Now showing 1 - 6 of 6
  • Thumbnail Image
    Item
    Advancing Cell Culture Engineering Through Mechanistic Model Optimization
    (2020-04) O'Brien, Conor
    Over the past few decades, the emergence of new classes of treatments, including protein therapeutics, gene therapies, and cell therapies, has ushered in a new era of medicine. Unlike small molecule therapeutics, these treatments are produced in or consist of cells, typically mammalian in origin. Processes have been developed to produce many of these drugs at large scale, often in stirred tank bioreactors. Significant effort has driven staggering increases in the productivity of these processes, enabling economical manufacturing, and the potential to drive down costs and make drugs more widely available. However, the bioreactor is not a natural environment for cells isolated from a multicellular mammalian organism. Many biological regulations are carried over from the cells’ origin and can result in numerous undesirable behaviors manifesting in the dense, highly productive reactor environment. In certain culture stages, or in the case of excess nutrient supply, cells will secrete undesirable metabolites including lactate, ammonia, and many byproducts of amino acid metabolism. These compounds can retard cell growth, or otherwise alter the potency or productivity of the cultures. Traditional biologics process development employs the use of statistical design of experiments, often encompassing several reactors run in parallel for multiple rounds of experiments over a few months. There is thus substantial room for improvement for both the outcome of the development process, such as an increase in titer, and the time it takes to complete the development stage. Given that cell culture processes share intrinsic similarities in their underlying mechanistic behavior, there exists significant opportunity to reduce the overall number of experiments needed for process development, scaling, and diagnostics using models rather than treating cell culture processes as a black box. In this thesis, we present the case for the use of mathematical optimization of mechanistic models to accurately describe cell culture processes and augment their behavior. We first outline recent advances in understanding of metabolic regulation and homeostasis. Cell signaling and metabolic networks interact over multiple time-scales and through multiple means, resulting in cell metabolism with nonlinear behavior that is consequently context-dependent. In the following sections of this work, we then develop an optimization framework which can efficiently be used for the design of experiments to rewire cellular metabolism through metabolic engineering, or to otherwise understand the biological requirements of different metabolic phenomena. This framework is first applied to the Warburg effect, a century-old unsolved problem of rapid lactate production in proliferating cells to identify which enzymes may be altered to mitigate the lactate production. This framework in then applied to the problem of hepatic gluconeogenesis to study metabolic disease. As the expression of the enzymes specific to gluconeogenesis is not sufficient for glucose production, we explore what other requirements exist for the synthesis of glucose from different substrates. The next portion discusses the construction and optimization of a bioprocess model which includes metabolism, signaling, cell growth, and the reactor environment. This model is fit to a manufacturing-scale dataset to explore the origins of process variability and potential mitigation strategies. In the final segment of this thesis, we explore another aspect of protein therapeutics: product quality. A model of N-glycosylation is optimized in conjunction with successive rounds of experimentation with the goal of improving the galactose content on an antibody. This work highlights the benefits of feeding back experimental data to refine model parameters for better design and prediction of subsequent experiments.
  • Thumbnail Image
    Item
    Camp Ripley Sentinel Landscape Climate Resilience Analysis and Strategic Plan Amendments
    (University of Minnesota Duluth, 2023-07) Bartsch, Will; Cai, Meijun; Johnson, Kris; Nixon, Kristi; Sprague, Tiffany; Wright, Chris; Olsen, Louis; Reed, Jane
    Camp Ripley is a military training facility located in central Minnesota. It is surrounded by the 750,000-acre Camp Ripley Sentinel Landscape (CRSL). Created in 2015, the CRSL consists of working and natural lands surrounding Camp Ripley with the purpose of protecting the training mission of the facility. The rural character of this landscape is generally compatible with that mission. However, it could be compromised by development, which could diminish habitat quality and raise the potential for conflict with landowners. The ability of Camp Ripley to maintain its mission is also threatened by a changing climate, which is projected to get warmer and wetter with a higher frequency of large precipitation events in the region. To help ensure the viability of the mission, the Natural Resources Research Institute assessed climate vulnerabilities and developed strategies to build and enhance climate resilience. Specifically, we 1) evaluated and selected Global Climate Models (GCM) that are expected to perform well in the region, 2) modeled stream water quantity and quality under different land use and climate scenarios, 3) characterized the landscape using Geographic Information Systems, 4) modeled and identified high-quality habitat for at-risk species, 5) evaluated and ranked parcels for conservation and restoration opportunities, 6) created afforestation plans for individual parcels, and 7) amended the Camp Ripley Strategic Plan with climate resilience language and strategies. Modeling stream quantity and quality under different land use scenarios indicates generally increased flow and sediment and nutrient concentration in scenarios where forest land is converted to agriculture or developed. Modeling under different future climate scenarios generally predicts decreased summer baseflow and increased nutrient and sediment concentrations. A suite of environmental data was acquired and developed to help characterize the landscape and prioritize parcels for conservation or restoration activity. Habitat models were developed for the Red-shouldered hawk, Golden-winged warbler, Northern long-eared bat, and Blanding’s turtle, all listed as at-risk or endangered under the Endangered Species Act. Afforestation plans with carbon sequestration modeling and carbon market participation compensation estimates were completed for two parcels within the landscape, illustrating an economically viable, market-driven solution. Climate resilience language was added to the strategic plan with emphasis placed on the restructuring and expansion of the strategy table while improving alignment with Minnesota’s Climate Adaptation Framework.
  • Thumbnail Image
    Item
    Modeling Outputs of Efficient Compressibility Estimators
    (2018-06) Asamoah Owusu, Dennis
    There are times when it is helpful to know whether data is compressible before expending computational resources to compress it. The standard deviation of the byte distribution of data is an example of a measure of compressibility that does not involve actually compressing the data. This work considered five such measures of compressibility: byte standard deviation, shannon entropy, “average meaning entropy”, “byte counting” and “heuristic method”. We developed models that relate the output of these measures to the compression ratios of gzip, lz4 and xz using data retrieved from browsing Facebook, Wikipedia and YouTube. The models for byte standard deviation, shannon entropy and “average meaning entropy” were linear in both the parameters and the variables. The model for “byte counting” was non-linear in the predictor variable but linear in the parameters. The “heuristic method” was a classification model. In general, there was a strong relationship between the measures and the compressibility of a given data. Also, in many cases the models developed using one set of data from a source (like Youtube) was able to estimate the compressibility of another data set from the same source to a useful extent. This suggests the potential for developing a model per ECE for a source that can predict, to a useful degree, the compressibility of data from that source. At the same time, the differences in accuracy when models were evaluated on the data they were developed from versus when evaluated on new data from the same source indicate that there are important differences in the nature of the data coming from even the same source.
  • Thumbnail Image
    Item
    A New Model for Acute Pain Management in Children: Examining Patient Characteristics and Potential Implications for Research and Practice
    (2021-05) Eull, Donna
    AbstractCurrent literature suggests that acute pain management in hospitalized children remains substandard, resulting in adverse physical, cognitive, and emotional effects for many children. Improvements to pediatric acute pain management require an updated conceptual model and validation of current assumptions from the literature. The purpose of the three studies in this dissertation was to advance the state of the science on acute pain management in hospitalized children through an updated conceptual model, a critical review of literature, and analysis of pain management data from a children's hospital. The new acute pain management model transforms the role of the nurse from gatekeeper to facilitator in genuine partnership with children and families. The critical review of the literature suggested that differences in medication type and frequency for acute pain are associated with children’s sex and race/ethnicity, however study limitations make it difficult to draw meaningful conclusions about potential disparities in acute pain management for children. Findings from a retrospective chart review suggested no differences by sex, race/ethnicity, or limited English proficiency (LEP) in the average number of pain assessments, medication by weight, or outcomes. The results from this study may indicate progress in the management of acute pain in hospitalized children, as children in this study demonstrated average pain intensity scores which typically indicate mild, well-controlled pain. Replication of this study in other hospitals is needed to determine organizational effects on pain outcomes. Future research should also focus on identifying the components needed to establish genuine partnerships with patients and families and the potential influence of parents on effective pain management.
  • Thumbnail Image
    Item
  • Thumbnail Image
    Item
    Real-Time System Identification and Control of Engine System Using Least Squares Learning and Simplex Tessellation
    (2022-12) Tranquillo, Holden
    To aid in engine control for achieving the stable combustion of varying cetane level fuels, a computationally efficient algorithm for the online learning of an engine model based on real-time input and output measurements is developed. Innovations in engine technology has led to the feasibility of robust, multi-fuel engine systems capable of operating on unknown or non-ideal fuel types. To attain such performance, advanced control strategies must be implemented in order to achieve stable engine combustion using such fuels. The method developed in this work, based on piecewise-linear modeling via discrete nodes and recursive linear least squares is first derived for the one-dimensional system of injection timing and combustion phasing. The learning model is then used for adaptive feedforward and feedback control of the SISO system in simulation using a gaussian process model as a virtual engine. The algorithm is then extended to the two-input/two-output system of injection timing and fuel mass and their effect on combustion phasing and indicated mean effective pressure (IMEP). Data generated using computational fluid mechanics is used to supplement experimental data in the development of the 2D model. The theory of barycentric and affine coordinates is introduced and applied to the concept of piecewise planes to approximate nonlinear surfaces. The learning model is utilized in an adaptive MIMO feedforward algorithm to control the engine to a desired combustion phasing and IMEP. Additionally, a decoupled integral feedback control scheme is presented and shown effective in simulation. A generalization of the learning algorithm for higher dimensions is made in order to model higher order systems. Specifically, simplex tessellation and barycentric coordinates as regressor coefficients are shown to generalize node locating and updating in arbitrary dimensions. The generalized learning algorithm is applied to a synthetic three-input data set in order show feasibility of the model for higher order nonlinear systems. The algorithm developed in this work is a unique, generalized, data-driven model capable of the real-time learning and control of multi-dimensional systems. The computational efficiency and generalization of the method allows for the real-time system identification of engine systems that are operating in unknown or untested environments. Existing engine models lack the efficiency to perform at the operating times seen in internal combustion engines. Implemented in a physical engine, the developed algorithm could be used for adaptive modeling of the system when undergoing a fuel or environmental change, which in turn can be used to aid in adaptive control of the engine. In commercial application, the real-time learning model could be used to decrease or eliminate the traditional in-house testing of engines required for lookup table generation, which would in turn decrease the time and cost in getting the engine to final application.