Matching Items (22)
Description
Arc Routing Problems (ARPs) are a type of routing problem that finds routes of minimum total cost covering the edges or arcs in a graph representing street or road networks. They find application in many essential services such as residential waste collection, winter gritting, and others. Being NP-hard, solutions are usually found using heuristic methods. This dissertation contributes to heuristics for ARP, with a focus on the Capacitated Arc Routing Problem (CARP) with additional constraints. In operations such as residential waste collection, vehicle breakdown disruptions occur frequently. A new variant Capacitated Arc Re-routing Problem for Vehicle Break-down (CARP-VB) is introduced to address the need to re-route using only remaining vehicles to avoid missing services. A new heuristic Probe is developed to solve CARP-VB. Experiments on benchmark instances show that Probe is better in reducing the makespan and hence effective in reducing delays and avoiding missing services.
In addition to total cost, operators are also interested in solutions that are attractive, that is, routes that are contiguous, compact, and non-overlapping to manage the work. Operators may not adopt a solution that is not attractive even if it is optimum. They are also interested in solutions that are balanced in workload to meet equity requirements. A new multi-objective memetic algorithm, MA-ABC is developed, that optimizes three objectives: Attractiveness, makespan, and total cost. On testing with benchmark instances, MA-ABC was found to be effective in providing attractive and balanced route solutions without affecting the total cost.
Changes in the problem specification such as demand and topology occurs frequently in business operations. Machine learning be applied to learn the distribution behind these changes and generate solutions quickly at time of inference. Splice is a machine learning framework for CARP that generates closer to optimum solutions quickly using a graph neural network and deep Q-learning. Splice can solve several variants of node and arc routing problems using the same architecture without any modification. Splice was trained and tested using randomly generated instances. Splice generated solutions faster that are also better in comparison to popular metaheuristics.
ContributorsRamamoorthy, Muhilan (Author) / Syrotiuk, Violet R. (Thesis advisor) / Forrest, Stephanie (Committee member) / Mirchandani, Pitu (Committee member) / Sen, Arunabha (Committee member) / Arizona State University (Publisher)
Created2022
Description
Coffee leaf rust (CLR) is an aggressive disease that has caused devastating production losses around the world. While coffee productivity under shade conditions has been simulated in sufficient detail, explicit modeling of the interactions between shading levels, microclimate, coffee plant physiology, and CLR progression remains largely unexplored, despite the recognized influence of shade on CLR epidemics. This dissertation introduces a new model, SpatialRust, as an initial approximation to an integrative simulation framework where farm design and management strategies can be linked to coffee production and CLR epidemic outcomes. SpatialRust considers stylized processes describing the dynamics of shade trees, coffee plants, and CLR and their interactions within a spatially explicit context. The dissertation then presents three experiments conducted using SpatialRust simulations. The first experiment investigates the role of shading as a mitigation tool for CLR outbreaks. It demonstrates that shade can effectively mitigate CLR epidemics when the conditions are otherwise conducive to a major CLR epidemic. Additionally, the experiment reveals complex effects of different shade management approaches, underscoring the potential value of future empirical studies that consider temporal and spatial shade dynamics in relation to CLR outcomes. The second experiment focuses on the financial balance of farms and examines how farmer preferences and needs influence farm management strategies. The findings indicate that incorporating CLR mitigation as part of the strategy's goals leads to positive long-term farm performance, even when planning for the short term. In the third experiment, the scope of the simulations is expanded to include neighboring coffee farms. The results demonstrate that the strategies adopted by immediate neighbors can affect the performance of individual farms, emphasizing the importance of considering the broader coffee-growing landscape context. This work shows that the integration of farm management practices and the resulting shading effects into a spatially explicit framework can provide valuable insights into the dynamics of CLR epidemics and how they respond to farmers' decisions.
ContributorsVanegas Ferro, Manuela (Author) / Janssen, Marcus A. (Thesis advisor) / Forrest, Stephanie (Committee member) / Restrepo Restrepo, Silvia (Committee member) / Arizona State University (Publisher)
Created2023
Description
The Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) pandemic, declared in March 2020 resulted in an unprecedented scientific effort that led to the deployment in less than a year of several vaccines to prevent severe disease, hospitalizations, and death from coronavirus disease 2019 (COVID-19). Most vaccine models focus on the production of neutralizing antibodies against the spike (S) to prevent infection. As the virus evolves, new variants emerge that evade neutralizing antibodies produced by natural infection and vaccination, while memory T cell responses are long-lasting and resilient to most of the changes found in variants of concern (VOC). Several lines of evidence support the study of T cell-mediated immunity in SARS-CoV-2 infections. First, T cell reactivity against SARS-CoV-2 is found in both (cluster of differentiation) CD4+ and CD8+ T cell compartments in asymptomatic, mild, and severe recovered COVID-19 patients. Second, an early and stronger CD8+ T cell response correlates with less severe COVID-19 disease [1-4]. Third, both CD4+ and CD8+ T cells that are reactive to SARS-CoV-2 viral antigens are found in healthy unexposed individuals suggesting that cross-reactive and conserved epitopes may be protective against infection.
The current study is focused on the T cell-mediated response, with special attention to conserved, non-spike-cross-reactive epitopes that may be protective against SARS-CoV-2. The first chapter reviews the importance of epitope prediction in understanding the T cell-mediated responses to a pathogen. The second chapter centers on the validation of SARS-CoV-2 CD8+ T cell predicted peptides to find conserved, immunodominant, and immunoprevalent epitopes that can be incorporated into the next generation of vaccines against severe COVID-19 disease. The third chapter explores pre-existing immunity to SARS-CoV-2 in a pre-pandemic cohort and finds two highly immunogenic epitopes that are conserved among human common cold coronaviruses (HCoVs). To end, the fourth chapter explores the concept of T cell receptor (TCR) cross-reactivity by isolating SARS-CoV-2-reactive TCRs to elucidate the mechanisms of cross-reactivity to SARS-CoV-2 and other human coronaviruses (HCoVs).
ContributorsCarmona, Jacqueline (Author) / Anderson, Karen S (Thesis advisor) / Lake, Douglas (Thesis advisor) / Maley, Carlo (Committee member) / Mangone, Marco (Committee member) / LaBaer, Joshua (Committee member) / Arizona State University (Publisher)
Created2023
Description
Graphic Processing Units (GPUs) have become a key enabler of the big-data revolution, functioning as defacto co-processors to accelerate large-scale computation. As the GPU programming stack and tool support have matured, the technology has alsobecome accessible to programmers. However, optimizing programs to run efficiently on GPUs requires developers to have both detailed understanding of the application logic and significant knowledge of parallel programming and GPU architectures.
This dissertation proposes GEVO, a tool for automatically tuning the performance of GPU kernels in the LLVM representation to meet desired criteria. GEVO uses population-based search to find edits to programs compiled to LLVM-IR which improves performance on desired criteria and retains required functionality. The evaluation of GEVO on the Rodinia benchmark suite demonstrates many runtime optimization techniques. One of the key insights is that semantic relaxation enables GEVO to
discover these optimizations that are usually prohibited by the compiler. GEVO also explores many other optimizations, including architecture- and application-specific. A follow-up evaluation of three bioinformatics applications at their different stages of development suggests that GEVO can optimize programs as well as human experts, sometimes even into a code base that is beyond a programmer’s reach. Furthermore, to unshackle the constraint of GEVO in optimizing neural network (NN) models,
GEVO-ML is proposed by extending the representation capability of GEVO, where NN models and the training/prediction process are uniformly represented in a single intermediate language. An evaluation of GEVO-ML shows that GEVO-ML can
optimize network models similar to how human developers improve model design, for example, by changing learning rates or pruning non-essential parameters. These results showcase the potential of automated program optimization tools to both reduce
the optimization burden for researchers and provide new insights for GPU experts.
ContributorsLiou, Jhe-Yu (Author) / Forrest, Stephanie (Thesis advisor) / Wu, Carole-Jean (Thesis advisor) / Lee, Yann-Hang (Committee member) / Weimer, Westley (Committee member) / Arizona State University (Publisher)
Created2023
Modeling Spatial Competition and Adaptive Therapy Protocols in Three-Dimensional Vascularized Tumors
Description
In contrast to traditional chemotherapy for cancer which fails to address tumor heterogeneity, raises patients’ levels of toxicity, and selects for drug-resistant cells, adaptive therapy applies ideas from cancer ecology in employing low-dose drugs to encourage competition between cancerous cells, reducing toxicity and potentially prolonging disease progression. Despite promising results in some clinical trials, optimizing adaptive therapy routines involves navigating a vast space of combina- torial possibilities, including the number of drugs, drug holiday duration, and drug dosages. Computational models can serve as precursors to efficiently explore this space, narrowing the scope of possibilities for in-vivo and in-vitro experiments which are time-consuming, expensive, and specific to tumor types. Among the existing modeling techniques, agent-based models are particularly suited for studying the spatial inter- actions critical to successful adaptive therapy. In this thesis, I introduce CancerSim, a three-dimensional agent-based model fully implemented in C++ that is designed to simulate tumorigenesis, angiogenesis, drug resistance, and resource competition within a tissue. Additionally, the model is equipped to assess the effectiveness of various adaptive therapy regimens. The thesis provides detailed insights into the biological motivation and calibration of different model parameters. Lastly, I propose a series of research questions and experiments for adaptive therapy that CancerSim can address in the pursuit of advancing cancer treatment strategies.
ContributorsShah, Sanjana Saurin (Author) / Daymude, Joshua J (Thesis advisor) / Forrest, Stephanie (Committee member) / Maley, Carlo C (Committee member) / Arizona State University (Publisher)
Created2023
Description
The innate immune system serves as an immediate response to pathogenic infection and an informant to the adaptive immune system. The 2′,5′-oligoadenylate (2-5A) synthetase (OAS)–RNase-L system is a component of the innate immune system induced by interferons (IFNs) and serves to eliminate viral infections. In humans, three enzymatically active OAS proteins exist, OAS1, OAS2, and OAS3. Recent evidence suggests variations in cellular localization of OAS proteins may influence the impact and influence of those proteins on viral replication. However, viral suppression mechanisms involving specific OAS proteins are still unclear for most viruses. Here, I overexpress different isoforms of OAS and determined that though viruses within the same family have similar replication strategies, the extent to which each OAS protein impacts viral replication for Flaviviruses, and Alphaviruses varies. In contrast to the innate immune system, the adaptive immune system provides specific and long-lived immune responses. In the context of cancer, T cells have been shown to play a prominent role in tumor regression. It has previously been demonstrated that administration α-CTLA-4/α-PD-L1 immune checkpoint blockade (ICB) to mice inoculated with a K7M2 metastatic osteosarcoma (mOS) cell line resulted in ~50% survival. Here, I sought to determine biological differences among murine responders and non-responders to ICB for mOS to understand better what factors could increase ICB efficacy. A prospective culprit is a variance in circulating antibodies (Abs). I have shown that sera from mice, before inoculation with mOS or ICB, display distinct differences in Ab repertoire between responders and non-responders, suggesting the presence or absence of particular Abs may influence the outcome of ICB. Recent studies have also shown that malleable environmental factors, such as differences in microbiome composition, can yield subsequent changes in circulating Abs.
Strong associations have been made between host-microbiome interactions and their effects on health. Here, I study potential associations of microbiome-mediated impacts on ICB efficacy for mOS. Additionally, I sought to determine potential changes in T-cellular response to mOS due to modulations in microbiome composition and showed that ICB efficacy can change in conjunction with microbiome composition changes in a murine model.
ContributorsDi Palma, Michelle Pina (Author) / Blattman, Joseph N (Thesis advisor) / Li, Yize (Thesis advisor) / Anderson, Karen S (Committee member) / McFadden, Grant (Committee member) / Arizona State University (Publisher)
Created2023
Description
Resistance to existing anti-cancer drugs poses a key challenge in the field of medical oncology, in that it results in the tumor not responding to treatment using the same medications to which it responded previously, leading to treatment failure. Adaptive therapy utilizes evolutionary principles of competitive suppression, leveraging competition between drug resistant and drug sensitive cells, to keep the population of drug resistant cells under control, thereby extending time to progression (TTP), relative to standard treatment using maximum tolerated dose (MTD). Development of adaptive therapy protocols is challenging, as it involves many parameters, and the number of parameters increase exponentially for each additional drug. Furthermore, the drugs could have a cytotoxic (killing cells directly), or a cytostatic (inhibiting cell division) mechanism of action, which could affect treatment outcome in important ways. I have implemented hybrid agent-based computational models to investigate adaptive therapy, using either a single drug (cytotoxic or cytostatic), or two drugs (cytotoxic or cytostatic), simulating three different adaptive therapy protocols for treatment using a single drug (dose modulation, intermittent, dose-skipping), and seven different treatment protocols for treatment using two drugs: three dose modulation (DM) protocols (DM Cocktail Tandem, DM Ping-Pong Alternate Every Cycle, DM Ping-Pong on Progression), and four fixed-dose (FD) protocols (FD Cocktail Intermittent, FD Ping-Pong Intermittent, FD Cocktail Dose-Skipping, FD Ping-Pong Dose-Skipping). The results indicate a Goldilocks level of drug exposure to be optimum, with both too little and too much drug having adverse effects. Adaptive therapy works best under conditions of strong cellular competition, such as high fitness costs, high replacement rates, or high turnover. Clonal competition is an important determinant of treatment outcome, and as such treatment using two drugs leads to more favorable outcome than treatment using a single drug. Switching drugs every treatment cycle (ping-pong) protocols work particularly well, as well as cocktail dose modulation, particularly when it is feasible to have a highly sensitive measurement of tumor burden. In general, overtreating seems to have adverse survival outcome, and triggering a treatment vacation, or stopping treatment sooner when the tumor is shrinking seems to work well.
ContributorsSaha, Kaushik (Author) / Maley, Carlo C (Thesis advisor) / Forrest, Stephanie (Committee member) / Anderson, Karen S (Committee member) / Cisneros, Luis H (Committee member) / Arizona State University (Publisher)
Created2023
Description
Nucleic acid nanotechnology is a field of nanoscale engineering where the sequences of deoxyribonucleicacid (DNA) and ribonucleic acid (RNA) molecules are carefully designed to create self–assembled nanostructures with higher spatial resolution than is available to top–down fabrication methods. In the 40 year history
of the field, the structures created have scaled from small tile–like structures constructed from a few hundred
individual nucleotides to micron–scale structures assembled from millions of nucleotides using the technique
of “DNA origami”. One of the key drivers of advancement in any modern engineering field is the parallel
development of software which facilitates the design of components and performs in silico simulation of the
target structure to determine its structural properties, dynamic behavior, and identify defects. For nucleic acid
nanotechnology, the design software CaDNAno and simulation software oxDNA are the most popular choices
for design and simulation, respectively. In this dissertation I will present my work on the oxDNA software
ecosystem, including an analysis toolkit, a web–based graphical interface, and a new molecular visualization
tool which doubles as a free–form design editor that covers some of the weaknesses of CaDNAno’s lattice–based design paradigm. Finally, as a demonstration of the utility of these new tools I show oxDNA simulation
and subsequent analysis of a nanoscale leaf–spring engine capable of converting chemical energy into dynamic motion. OxDNA simulations were used to investigate the effects of design choices on the behavior of
the system and rationalize experimental results.
ContributorsPoppleton, Erik (Author) / Sulc, Petr (Thesis advisor) / Yan, Hao (Committee member) / Forrest, Stephanie (Committee member) / Stephanopoulos, Nicholas (Committee member) / Arizona State University (Publisher)
Created2022
Description
Emerging pathogens present several challenges to medical diagnostics. Primarily, the exponential spread of a novel pathogen through naïve populations require a rapid and overwhelming diagnostic response at the site of outbreak. While point-of-care (PoC) platforms have been developed for detection of antigens, serologic responses, and pathogenic genomes, only nucleic acid diagnostics currently have the potential to be developed and manufactured within weeks of an outbreak owing to the speed of next-generation sequencing and custom DNA synthesis. Among nucleic acid diagnostics, isothermal amplification strategies are uniquely suited for PoC implementation due to their simple instrumentation and lack of thermocycling requirement. Unfortunately, isothermal strategies are currently prone to spurious nonspecific amplification, hindering their specificity and necessitating extensive empirical design pipelines that are both time and resource intensive. In this work, isothermal amplification strategies are extensively compared for their feasibility of implementation in outbreak response scenarios. One such technology, Loop-mediated Amplification (LAMP), is identified as having high-potential for rapid development and PoC deployment. Various approaches to abrogating nonspecific amplification are described including a novel in silico design tool based on coarse-grained simulation of interactions between thermophilic DNA polymerase and DNA strands in isothermal reaction conditions. Nonspecific amplification is shown to be due to stabilization of primer secondary structures by high concentrations of Bst DNA polymerase and a mechanism of micro-complement-mediated cross-priming is demonstrated as causal via nanopore sequencing of nonspecific reaction products. The resulting computational model predicts primer set background in 64% of 67 test assays and its usefulness is illustrated further by determining problematic primers in a West Nile Virus-specific LAMP primer set and optimizing primer 3’ nucleotides to eliminate micro-complements within the reaction, resulting in inhibition of background accumulation. Finally, the emergence of Orthopox monkeypox (MPXV) as a recurring threat is discussed and SimCycle is utilized to develop a novel technique for clade-specific discrimination of MPXV based on bridging viral genomic rearrangements (Bridging LAMP). Bridging LAMP is implemented in a 4-plex microfluidic format and demonstrates 100% sensitivity in detection of 100 copies of viral lysates and 45 crude MPXV-positive patient samples collected during the 2022 Clade IIb outbreak.
ContributorsKnappenberger, Mark Daniel (Author) / Anderson, Karen S (Thesis advisor) / LaBaer, Joshua (Committee member) / Roberson, Robert (Committee member) / Lindsay, Stuart (Committee member) / Arizona State University (Publisher)
Created2023
Description
Nations censor specific information in accordance with their political, legal, and cultural standards. Each country adopts unique approaches and regulations for censorship, whether it involves moderating online content or prohibiting protests. This paper seeks to study the underlying motivations for the disparate behaviors exhibited by authorities and individuals. To achieve this, we develop a mathematical model designed to understand the dynamics between authority figures and individuals, analyzing their behaviors under various conditions. We argue that individuals essentially act in three phases - compliance, self-censoring, and defiance when faced with different situations under their own desires and the authority's parameters. We substantiate our findings by conducting different simulations on the model and visualizing the outcomes. Through these simulations, we realize why individuals exhibit behaviors falling into one of three categories, who are influenced by factors such as the level of surveillance imposed by the authority, the severity of punishments, the tolerance for dissent, or the individuals' boldness. This also helped us to understand why certain populations in a country exhibit defiance, self-censoring behavior, or compliance as they interact with each other and behave under specific conditions within a small network world.
ContributorsNahar, Anish Ashish (Author) / Daymude, Joshua (Thesis director) / Forrest, Stephanie (Committee member) / Barrett, The Honors College (Contributor) / Computer Science and Engineering Program (Contributor)
Created2024-05