21793 research outputs found
Sort by
Elucidating The Structure and Dynamics of Aurein Peptides in Solution and Micellar Environments Using Biomolecular Simulations
No full textAntimicrobial peptides (AMPs) are part of the innate immune system of many species from plants to mammals, showing antimicrobial activity against a wide range of microorganisms. Aurein peptides are a family of AMPs isolated from amphibians, displaying antibacterial and anticancer activity. In this work, we study the structure and dynamics of a set of Aurein peptides in aqueous solution and micellar environment using atomistic molecular dynamics (MD) simulations. We study six Aurein peptides, four of which are experimentally known to have antibacterial and anticancer activity (Aureins 1.2, 2.2, 2.5, and 3.1); and two longer Aurein peptides (Aureins 4.1 and 5.1) that are inactive. We build Markov state models (MSMs) to investigate the conformations of the peptides and transitions between different folding intermediates. Aurein 1.2 is an experimentally well-characterized peptide; there is literature reporting on the structures of Aurein 2.2 and 2.5. Our data show that the active Aureins visit three conformational states in solution: alpha-helical, partially alpha-helical, and random coil. The inactive Aureins show rapid transitions between short-lived conformations, and their MSMs were not reproducible. The simulations of Aureins 1.2 and 2.5 in micelles has shown that these peptides retain their folded α-helical structure in a hydrophobic environment. The secondary amide proton chemical shifts of Aurein 1.2 in micelle environments differing in the number of detergent molecules were computed and compared to the available experimental data. The micelle simulations comprising 80 SDS molecules agree well with the experimental data, suggesting that our simulations can predict the folding of Aurein peptides in micelles. Our results show that combining molecular simulations with MSMs enables the prediction of in-micelle conformations for Aurein peptides lacking experimental structures
Sport Specialization and the Association with History of Injury in Young Iranian Athletes
Get PDFPurpose: Sport specialization can cause physical injuries in athletes, but the impact of sport specialization in child and adolescent athletes aged 8 to 18 in Iran is unknown. This study aims to examine the presence of sport specialization and its associations with sex, sport type, and history of sport-related injuries.
Method: Participants completed two questionnaires. The first questionnaire consisted of the sports specialization level and participants’ demographic data, including age, sex, sport type, and physical activity participation (hours of sport participation, rest days). The second questionnaire focused on participants’ sport-related injury history.
Result: The study confirmed the presence of youth sport specialization among athletes in this age group in Iran. No association was found between injury history and participants' sex, sport type, or sport specialization.
Conclusion: While we found the presence of sport specialization in Iran, there was no significant association between participants' sport specialization, sex, sport type, and history of sport-related injuries. However, athletes that trained more than 8 months annually in a primary sport had a positive association with prior injuries. There was a significant difference in training intensity, particularly when exceeding 8 months per year, between those with and without a history of injury. This difference underscores the impact of extended training on injury susceptibility
Classification of Breast Cancer Cytological Images using Vision Transformers
Get PDFThis thesis evaluates the effectiveness of Vision Transformers (ViT) and Swin Transformers for breast cancer classification, highlighting their advantages over traditional Convolutional Neural Networks (CNNs) in processing cytological images. Amid the critical need for better breast cancer diagnostics, these transformer-based models emerge as a promising solution, adept at capturing complex spatial and contextual data in medical images.
The research methodology involved collecting and preprocessing a dataset of cytological and histopathological breast cancer images. The performance of the vision transformers was assessed using metrics such as accuracy, precision, recall, and AUC-ROC, and compared against established CNN architectures. The results demonstrate that vision transformers excel at extracting complex patterns from images, significantly outperforming current methods. Specifically, the study reports a 3.06% improvement in classification accuracy over traditional approaches, achieving 95.01% accuracy on test sets and perfect accuracy in validation.
The thesis underscores the potential of ViT and Swin models to advance early detection and diagnosis of breast cancer. Their success in the study suggests a transformative shift towards utilizing advanced deep learning architectures in medical image analysis. This approach not only enhances diagnostic accuracy but also offers a data-efficient solution to the challenges of breast cancer classification. The findings advocate for further exploration of transformer-based models, which could redefine the standards of computer-aided diagnosis and significantly impact the field of cancer classification
Supplementary Appendix: Data tables for: "Scaling Up Upcycling: A Comparative Analysis of Furniture, Lighting and Bags Made Through Repurposing"
Get PDFAppendix 1: Data tables for the article "Scaling Up Upcycling: A Comparative Analysis of Furniture, Lighting and Bags Made Through Repurposing"
Appendix 1A: Seating
Appendix 1B: Small Tables
Appendix 1C: Lighting
Appendix 1B: Bag
Unsupervised Learning Based on Multivariate Libby-Novick Beta Mixture Model for Medical Data Analysis
Get PDFThis thesis proposes a set of innovative clustering techniques that lever- age finite and infinite mixture models to analyze medical data and images of cells. The proposed approaches are designed to improve the accuracy and efficiency of clustering in these domains. These models utilize a flexible distribution, the Libby-Novick Beta distribution, to better model data with varying shapes due to an additional shape parameter compared to the con- ventional Beta distribution. In this study, our initial approach involves the use of deterministic learning techniques, with a focus on maximum likelihood using the expectation-maximization approach. To achieve accurate data rep- resentation in unsupervised learning, it is crucial to determine the optimal number of clusters. So, we expand the minimum message length (MML) principle to ascertain the number of clusters in Libby-Novick Beta mixtures. In order to overcome the challenge of estimating the number of mixture components, we extend our finite mixture model to an infinite one. Nonparametric Bayesian techniques can effectively capture data distribution with an unknown number of components. This approach is useful for complex data sets and can lead to more accurate predictions and better decision-making. Our models are evaluated for different medical applications throughout the entire process, and they consistently show superior performance over traditional alternatives. This study reveals the significance of the Libby-Novick Beta distribution and the recommended mixture models in converting medical data into practical insights. This conversion aids healthcare professionals in making more accurate decisions, thereby advancing the overall healthcare field
Teklė and the Women
No full textFor this thesis, I have written a series of interconnected short stories that tell the early life of Teklė, a Lithuanian immigrant to Canada in the 1930s. These stories are written alternately from Teklė’s point of view as well as that of different women who come in and out of her life. In this work, I aim to paint a picture of the ways women recognize and connect with one another through common experiences, in turn telling the story of Teklė’s hardships, unwavering hope, and determination to find family and place. Teklė is based loosely on my own grandmother, and each story in the collection centres around people and events that influenced her life.
In the women’s stories, I explore themes of grief, caretaking, marital conflict, child-rearing, and other responsibilities carried by women across generations. Each woman is compelled by Teklė’s presence to reflect on their own life, and in turn help Teklė to heal and move on.
Interspersed are stories told from Teklė’s point of view, providing an opportunity to see her unfiltered at pivotal moments. Although the themes of abuse, scarcity, discrimination, and sexism run through the collection, Teklė remains a character with unshakable hope. Hers is ultimately the story of the female immigrant experience, where new lives are built on a foundation of trauma and tragedy that’s often walled away for survival. This collection is an attempt to remove some of the bricks for Teklė
Investigation of Ionic-Liquid-Based Dispersive Liquid-Liquid Microextraction and Lipid Adduct Formation for Untargeted Lipidomics
No full textLipidomics is the study of lipids in biological matrices. Challenges in the field include the use of toxic solvents and data harmonization. Ionic liquids (ILs), organic salts that can extract organic compounds and biomolecules, offer potential solutions, has not been extensively studied for lipidomics. Two ILs, 1-hexyl-3-methylimidazolium hexafluorophosphate and trihexyltetradecylphosphonium hexafluorophosphate, were assessed for lipid extraction using IL dispersive liquid-liquid microextraction with methanol as a disperser. Both ILs showed selectivity towards lipids of intermediate polarity, with limited extraction efficiency towards more polar lipids. However, slight solubility of ILs in the aqueous layer led to significant ion suppression for lipids tested when analyzed using liquid chromatography-mass spectrometry. Overall, the tested ILs were unsuitable for untargeted lipidomics of plasma.
Internal standards are commonly used for semi-quantification in untargeted lipidomics. However, electrospray ionization (ESI) differences between internal standards and endogenous lipids can impact quantification accuracy. The adduct formation for phospholipids, glycerolipids, sphingolipids, and sterol lipids was studied in human plasma, murine plasma, murine spinal cord, and murine liver tissue. Within-class adduct formation was consistent for most lipid classes, except for diglyceride and cholesteryl ester lipid subclasses, necessitating multiple internal standards. Mismatch between internal standard and endogenous lipid adduct formation was observed for phosphocholine, ceramide, triglyceride, diglycerol, and cholesteryl ester lipid subclasses in at least one of the four matrices. Inter-batch variability in sodium concentrations significantly influenced sodium adduct formation and could not be controlled with usage of ammonium volatile salts. Finally, monoglycerides and diglycerides were detected to form acetate adducts in ESI-
Jarring Lots: The Road Toward Decolonizing Polish Exhibitions will not be a Straight One
No full textThe origins of postcolonial discourses in Poland today are connected to a prevalent feeling among Poles that Poland itself has been an ‘overlooked’ victim of colonialism. In relation to the contemporary discourse on ’care and repair,’ this research-creation engages with the emotional and epistemological dilemmas inherent to showcasing Poland’s entanglement with the ‘global colonial power matrix’. In this written component of my thesis, I present a review of current practices within Poland’s art scenes to illustrate how coloniality is both reproduced and challenged in Polish museums. Subsequently, I reflect on my creative process behind producing the exhibition Jarring Lots. Drawing on Édouard Glissant, the name of the exhibition is a linguistic and conceptual reference to the fermentation of dill pickles and the transformative potential of opacit. Ultimately, this research-creation proposes that what is lacking in Poland’s exhibitions are decolonial gestures focused on taking care of the country’s racialised communities
Mantab Navigating Through a Global Disruption
No full textThis paper will seek to examine as to how Mantab, a frozen fruit and vegetable importer based in Eastern Canada overcame the challenges brought upon by the Covid-19 pandemic. This case study will identify Mantab’s identification, prevention, and mitigation strategies to each risk the organization is exposed to both internally and externally. Additionally, an in-depth investigation into Mantab’s financial health will be conducted and how it contributed to the decision-making processes made during the pandemic. Lastly, an investigation into the company’s risk tolerance and risk appetite level will be analyzed and determining whether they are in fact justifiable. The purpose of this study is to understand which decisions resulted in a positive or negative outcome and if they led to the organization gaining or losing market share. In retrospect, this will enable the organization to strengthen their overall operational resiliency and provide better insight on how to navigate through future turbulent working environments
Machine Learning for Anomalies Detection in Real-time Cloud
Get PDFCloud computing enables on-demand access to shared resources hosted in data centers and managed by cloud service providers. However, as cloud environments scale in size and complexity, they become increasingly prone to anomalies—deviations from expected behavior—that can disrupt reliability and availability. In real-time clouds, where operations must be completed within strict time constraints, anomalies pose a greater risk, potentially causing cascading failures, degraded performance, and increased maintenance costs. To address these issues, efficient methods for anomaly detection in real-time cloud environments are essential for maintaining service quality and operational efficiency.
Machine Learning (ML) has emerged as a promising approach for detecting anomalies in real-time clouds. By analyzing high-dimensional data such as system logs, traces, and performance metrics, ML models can identify complex patterns and deviations in dynamic, large-scale, and heterogeneous environments. However, employing ML in real-time clouds introduces several challenges, including handling sequential performance metrics, where evolving system behaviors cause concept drift, degrading model accuracy and requiring rapid adaptation to maintain low-latency anomaly detection; analyzing distributed traces, where inter-service dependencies and dynamic workloads introduce latency-sensitive bottlenecks, making timely anomaly detection difficult; and detecting anomalies in contextual logs, where log instability, class imbalance, and labeling dependency hinder model learning, further complicating real-time anomaly response under strict time constraints.
This thesis addresses these challenges with three key contributions for real-time cloud environments, where low-latency, adaptive, and scalable anomaly detection is critical. First, we propose a concept drift adaptation algorithm that integrates prediction-driven anomaly detection and adaptive window-based methods. This approach ensures effective handling of concept drift by dynamically adjusting to changes in the data distribution, enhancing detection accuracy over time. Second, we introduce a graph-based learning approach that captures inter-service dependencies while leveraging collaborative learning to reduce computational overhead and enable real-time updates.
Third, we present a self-supervised log anomaly detection that adapts to evolving log structures without requiring labeled data, improving detection efficiency in dynamic cloud environments