82837 research outputs found
Sort by
Identification of Key Genes and Pathways in Ovarian Cancer Using Weighted Gene Co-expression Network Analysis
No full textBackground: Ovarian cancer is one of the most lethal gynecological malignancies, largely due to late-stage diagnosis, high tumor heterogeneity, and limited treatment options. Understanding the molecular mechanisms underlying its progression is critical for improving early detection and identifying new therapeutic targets.
Objective: This study aimed to identify key genes, biological pathways, and regulatory microRNAs associated with ovarian cancer by analyzing gene expression profiles using a systems biology approach.
Methods: Gene expression data from the GSE38666 dataset were retrieved and analyzed. Differentially expressed genes (DEGs) between ovarian tumor and normal tissues were identified using the limma package in R. A weighted gene co-expression network was constructed using WGCNA, and modules significantly correlated with tumor stage were examined. Hub genes were identified using topological measures including Degree, EPC, MNC, and EcCentricity. Functional enrichment analysis was conducted using GO and KEGG databases. Additionally, microRNA–mRNA interactions were explored to highlight post-transcriptional regulation of hub genes.
Results: The blue module exhibited the strongest positive correlation with tumor stage. Within this module, hub genes such as CDK1, CCNB1, and GAPDH were identified. These genes are involved in essential cancer-related pathways, including cell cycle regulation and the p53 signaling pathway. Several regulatory microRNAs including hsa-miR-145-3p, hsa-miR-299-3p, hsa-miR-429, and hsa-miR-630 were also identified as potential upstream regulators of CDK1 and CCNB1, suggesting their role in chemoresistance and tumor progression.
Conclusion: This study identifies key genes and regulatory microRNAs that may serve as diagnostic biomarkers or therapeutic targets in ovarian cancer. The integration of WGCNA and microRNA–mRNA network analysis provides a comprehensive understanding of the molecular landscape of ovarian cancer and lays the groundwork for future experimental validation and clinical application
Indigenous Community Responses to Wildfire and Climate Change: Mental Health and Wildland Firefighters
No full textSSHRC IDG awarded 2025: Wildfires in Alberta are becoming larger and more destructive due to climate change, with major events affecting entire communities and drawing international attention. However, the disproportionate impact on Indigenous communities and Indigenous wildland firefighters has received limited attention in both media and research. Indigenous communities—especially in Northern Alberta—are more vulnerable to wildfires and often face severe consequences, including loss of homes and infrastructure, evacuations, environmental damage, disruption of cultural and land-based activities, and significant mental health challenges. At the same time, Indigenous firefighters make up a large portion of Alberta’s wildland firefighting workforce and are frequently responsible for protecting their own communities, adding emotional strain to already dangerous and demanding work. Despite extensive research on wildfire impacts in non-Indigenous populations, little focus has been placed on the long-term and cumulative effects of worsening wildfire seasons on Indigenous communities, including the mental health of both residents and firefighters. Existing studies tend to center on single evacuation events rather than ongoing climate-driven wildfire threats. This research proposes a collaborative, participatory approach with Indigenous communities to better understand their experiences, resilience, and mental health challenges related to worsening wildfires. It aims to fill critical gaps in the literature, highlight the disproportionate burdens Indigenous communities face, and bring greater public awareness to these overlooked impacts
Female Mallard
No full textThis female mallard was walking on the trail at the Clifford E. Lee Nature Sanctuary. Mallards are one of the most common and numerous waterfowl at the Sanctuary
"It's a sign of deep love and connection": Honouring ecological grief to support wellbeing and mobilize climate action
No full textAs the ongoing burning of fossil fuels continues to drive ecological destabilization, grief for ecological loss and its wide-ranging impacts on personal and collective life is widespread. Researchers agree that feeling impacted by the tremendous losses produced by climate change and its associated injustices is a reasonable and adaptive response. Bereavement studies have long understood grief as an expression of love and attachment for that which is lost. Similarly, climate grief may connect powerful personal feelings with collective experiences of loss and vulnerability caused by climate change, revealing personal values and shared interests that may inspire climate action. However, the absence of collective practices for honoring ecological grief leads to its “disenfranchisement” by depriving grievers of social scripts, rituals, and community support, and undermining well-being. Despite the prevalence of ecological grief and its generative potential, research investigating the relationship between climate grief, well-being, and engagement with the climate crisis—as well as the impact of interventions that collectively honour and enfranchise climate grief—is limited.
To address this gap, this thesis research had four objectives: 1) Characterize the lived-experience of ecological and climate grief; 2) Explore how ecological and climate grief connected to personal values; 3) Examine how ecological and climate grief shaped engagement with the climate crisis; and 4) Investigate the impact of collective practices for honouring and processing climate grief on wellbeing and capacity for climate action.
Semi-structured, in-depth interviews were conducted with adults (n = 15) who had attended Refugia Retreats programs in Alberta, Canada where they had engaged in facilitated group practices for sharing and processing climate grief. Interviews were analyzed using reflexive thematic analysis. Results demonstrate that participants’ experiences of ecological grief were indeed fertile sites for ecological, social, and political consciousness raising. Interviewees’ grief revealed interconnections between ecological, personal, and collective vulnerabilities caused by the climate crisis, leading to empathetic concern and solidarity with both human and more-than-human communities. Framing difficult emotions about climate change as grief supported interviewees in destigmatizing their feelings by foregrounding the larger socio-ecological conditions to which their feelings responded while also connecting them to the love and values that underpin difficult emotions. Witnessing both their own and other’s grief transformed some of their heaviest emotional burdens and opened up access to positive experiences such as love, joy, and agency. Because climate change often shaped other forms of loss and difficulty, collective practices for processing climate grief also helped interviewees cope with personal loss. Collectively honouring climate grief enhanced interviewees capacity for positive emotions associated with wellbeing and climate action while also allowing them to practice forms of collectivity that support collective action.
This research demonstrated how climate grief connected people with not only the personal, ecological, and collective toll of the climate crisis, but also to their deepest values and longings. Collective practices for honouring climate grief enabled participants to holistically process difficult emotions in ways that contributed to both individual emotional resilience, and capacity for collaborative and sustained action. This thesis contributes to our understanding of the generative potential of ecological grief as a leverage emotion that significantly shapes wellbeing and capacity for meaningful and transformative climate action
An Institutional Ethnography of Community Mental Health Care for Patients Facing Structural Barriers
No full textNorth America is grappling with escalating mental health crises characterized by rising mental illness prevalence, inequitable access to care, and severe workforce shortages. Recent discourse emphasizes the need to integrate social and structural determinants of health—such as poverty, violence, houselessness, and discrimination—into health care service design and delivery. Theoretical frameworks such as intersectionality and structural competency help to conceptualize the additive and exacerbating effects of institutionalized forms of discrimination on mental illness and recovery. However, the materiality of structural barriers– the practices that organize and coordinate these realities – remain underexplored.
Using an institutional ethnographic (IE) approach, this dissertation explores the social organization of community mental health care for patients living with serious mental illness inside conditions of poverty, violence, houselessness, and discrimination – the concrete realities that are abstracted into the theoretical structural competency framework. The specific objectives of this work were to: 1) critically examine circulating discourses in mental health as related to conceptualizations of structural competency; 2) empirically describe the work of caring for patients facing structural barriers; 3) and map the social relations of providing mental health care in the community. I examined the everyday work practices of Psychiatric Mental Health Nurse Practitioners (PMHNPs) to uncover the ruling relations that govern their work with patients with serious mental illness who live within conditions of poverty, violence, houselessness, and discrimination. Nine PMHNPs from outpatient community mental health clinics in a large California city participated in the study. Data collection included in-depth interviews, clinic observations, and analysis of relevant clinical and organizational texts. Data analysis involved IE techniques of indexing, mapping, and writing accounts.
This dissertation contains a concept analysis of structural competency completed pre-candidacy, which I included to demonstrate the evolution of my approach to this topic over the course of my doctoral journey (Chapter 2). I also wrote three manuscripts detailing IE findings. First, I examined how powerful circulating discourses about “access to care” were rhetorically aligned with institutional goals but practically misaligned with the everyday realities faced by PMHNPs and their patients (Chapter 3). Second, I explicated how the institutional emphasis on efficiency and billing, as mediated through the electronic health record (EHR), often organizes PMHNPs to prioritize diagnostic and medication management over addressing the broader social needs of patients (Chapter 4). Third, I used lessons from conducting this dissertation research to offer methodological insights into using IE for novice researchers, and to make the case for using IE to contribute useful qualitative knowledge in service of addressing complex public health issues (Chapter 5).
To improve the support and sustainability of PMHNPs’ work as frontline providers in community mental health care who have expert knowledge of patient wants and needs, practice coordinators and policymakers should consider reforms that integrate the knowledge and experiences of PMHNPs and their patients. Recommendations include: revising billing and reimbursement structures to recognize the full spectrum of care, including non-medical aspects; implementing flexible and contextually sensitive scheduling to accommodate patients with chronic mental illness; allowing for walk-in appointments and non-tech-mediated communication; adopting collaborative care models involving PMHNPs, social workers, and peer advocates to address clinical and social determinants of health; and realigning clinical documentation to meet the holistic needs of patients. Future research should employ institutional ethnography as well as other qualitative and participatory action methods to further explore mental health care practices and evaluate the impact of policy changes, ensuring the voices of affected individuals are prioritized
Fault Detection and Recovery: From Slow Feature Analytics to Operator-in-the-Loop Machine Learning
No full textIn today’s industrial processes, it is normal to measure and store thousands of interconnected process variables. To make sense of this data, dimensionality reduction techniques are frequently used to create useful features by eliminating redundant information. One approach, called slow feature analysis, identifies underlying data patterns that change gradually over time. A probabilistic extension of this method
was introduced to handle data in the presence of measurement noise. However, industrial process data presents further challenges beyond noise alone, including multi-
operating regions or multi-operating modes, the presence of outliers in the process measurements, non-stationarity due to aging equipment, and process drift as a result of process changes. When applying a basic probabilistic slow feature model to complex industrial data, the estimated parameter variance tends to be large. Therefore, this thesis aims to improve the probabilistic slow feature model to better handle the various complexities found in industrial settings.
Nonlinearity in measured data often arises due to the involvement of various complex operating units in industrial processes. Additionally, different product demands
and safety regulations necessitate multiple operating modes. The presence of outliers in the data hinders both standard slow feature analysis and its probabilistic variant
from effectively capturing data patterns. This limitation of conventional probabilistic slow feature analysis in handling these problems is largely due to constraints in
the state-transition matrix structure and the assumptions made about the data distributions. To address this challenge, this thesis introduces a robust multi-mode probabilistic slow feature analysis model. This model relaxes the assumptions about data distribution and incorporates multiple switching conditional random fields to account for nonlinearity. Furthermore, the distinction between developing models
from a plant-wide versus a unit-specific perspective highlights the need for tailored solutions. As a second contribution, this thesis proposes a moving window weighted probabilistic slow feature analysis model to handle unit-specific operations with near linear behavior. For plant-wide modeling, a slow feature analysis-based variational autoencoder is introduced to offer a more comprehensive solution.
High-dimensional datasets are often rooted in a low-dimensional latent space, and not all latent features influence every quality variable. Therefore, it is important to ensure that only a subset of latent variables impacts each quality variable. Additionally, experienced operators accumulate valuable insights through their operations, making their knowledge and feedback essential for fault identification. The third
contribution of this work introduces a reservoir computing-based slow feature analysis model, designed to extract slowly varying patterns from highly nonlinear process
data. A majority voting ensemble algorithm is employed to uncover the causal relationships between latent features. Following this, a graph neural network is trained
using the extracted features and the causal map generated by the previous models to classify faults. Real-time operator feedback is then incorporated to update the
classifier, enabling it to handle previously unseen faults. In this contribution, operator expertise is only leveraged for model updates and fault identification, and is not
directly applied during model training or parameter estimation.
Recognizing the growing impact of large language models and transformers in predictive modeling, the fourth contribution explores their potential in fault detection.
An operator-in-the-loop Bayesian variational transformer model is developed, where operator knowledge is directly integrated into the training process. This trained
model is then utilized for fault detection, bridging operator experience with advanced machine learning techniques. Once a fault is identified and if it is clear the process is
operating in suboptimal conditions, it becomes desirable to restore the system to its normality or even to its optimal performance. While traditional optimization meth-
ods are well-established, they often overlook the value of operator insights. Operator feedback, in the form of defining optimal and suboptimal regions and suggesting the
next sampling points, can significantly enhance this process. The fifth contribution proposes an operator-in-the-loop Bayesian optimization framework, which integrates
operator feedback into standard Bayesian optimization to achieve more effective process recovery and optimization.
The effectiveness of all contributions is demonstrated through simulations and industrial/experimental case studies, where they are compared to numerous state-of-the-art methods. These comparisons yield evidence of their effectiveness
Nonlinear motion control of single and multi aerial vehicles carrying a slung payload
No full textThis thesis addresses the motion control problem of single and multi unmanned aerial vehicles (UAVs) transporting a slung load using cables. For the single UAV case, this setup is called a slung load system(SLS). The SLS design has the advantage of ensuring UAV maneuverability while performing a payload transportation task. It also keeps the payload at a safe distance from the fast-spinning propellers and the onboard battery. In addition, the SLS design can be extended to multiple UAVs working cooperatively to carry and transport a shared slung payload. This cooperation enhances payload lift capacity and injects mechanical redundancy, which during unexpected faults, can be used to restore a nominal flight. SLSs possess rich and underactuated dynamics, hence, posing interesting yet challenging motion control problems. Precise payload motion control is a fundamental requirement for successful aerial transportation tasks. Motivated by this application, this thesis presents several control designs, ensuring precise payload transportation. Each control design defines a distinct control objective. For the multi-SLS (MSLS) setup, inter-UAV distances must be controlled to avoid dangerous collisions. Therefore, the motion control problem imposed by MSLSs is even more challenging as compared to a single SLS problem. This thesis also addresses the motion control problem of MSLSs.For a single SLS, trajectory-tracking and path-following controllers are developed for payload position control; both with exponential stability properties on a wide range of the state space. UAV heading angle is also directly controlled. Differential flatness and input-output exact feedback linearization are key tools. The linearization is achieved either using a dynamic or a quasi-static controller. Also, using a nested-loop control structure, the thesis develops a geometric path-following controller allowing for acrobatic manoeuvres. In this approach, three control loops are nested where linearization occurs only at the outermost loop.For an MSLS, the thesis proposes a novel flat output, which is used to design a dynamic input-output linearizing feedback. Besides payload motion control, this feedback also precisely controls the inter-UAV distances to avoid collisions during flight. The feedback offers exponential stability on a wide range of the state space
Mechanisms of sodium homeostasis in 3 species of freshwater fish exposed to pH 4.0 water
Get PDFThe uptake and efflux of ions in freshwater environments by teleost fish gills is an important regulatory process to maintain ionic concentration and acid-base balance in the gill cells. Freshwater environments pose a challenge on ionoregulatory processes due to the low environmental concentration of Na+ and other ions. Molecular transmembrane transporters like Na+/H+ exchangers help mitigate Na+ loss and maintain ionoregulation. However, in low pH freshwater conditions, circumneutral-acting molecular transports are inhibited due to unfavoured thermodynamic function.
There have been several proposed molecular mechanisms involved in the Na+ homeostatic response to low pH freshwater environments. However, identifying these mechanisms in a particular species is challenging due to a number of common caveats in molecular models. These include the limited range of functional pHs of a particular transporter, expression across a wide range of teleosts and limitations in function by thermodynamics. Recently, K+-driven Na+ uptake mechanisms and patterns of Na+ recovery in low pH freshwater environments have been proposed in zebrafish (Danio rerio) (Clifford et al. 2022). However, no mechanism was experimentally defined and the pattern of Na+ homeostasis in low pH freshwater has not been established for other species.
The goal of my thesis is to investigate patterns of Na+ homeostasis in multiple low pH-tolerant teleost fish exposed to low pH freshwater using 22Na+ unidirectional flux experiments. Zebrafish Na+ uptake was found to increase in parallel with increased K+ efflux after 8 h of pH 4.0 exposure, confirming the findings of Clifford et al. 2022. However, fathead minnows and yellow perch had different patterns of Na+ homeostasis in low pH freshwater. Fathead minnows displayed no differential Na+ homeostatic response in pH 4.0 water while yellow perch had an initial increased net Na+ efflux that recovered by 8 h within pH 4.0 exposure. Yellow perch Na+ uptake was also amiloride insensitive. This suggests that fathead minnows and yellow perch have different molecular mechanisms that result in different responses to low pH conditions, while still providing low pH tolerance.
I investigated candidate genes involved in zebrafish Na+ homeostatic response to low pH freshwater using differential expression of gene (DEG) analysis and validation through quantitative polymerase chain reaction (qPCR) analysis. Synaptic vesicle 2 protein like (svopl) was found to be significantly upregulated and have increased expression in zebrafish gill tissue exposed to 8 h of pH 4.0 water. Svopl has predicted functionality as an ion transporter with 12 transmembrane domains but experimental studies and information on Svopl and its human orthologues are extremely limited. I have identified Svopl as a valid candidate gene for the Na+ homeostatic response in zebrafish to low pH freshwater
Setting and dynamics of recent permafrost landslides in the central Mackenzie Valley, Northwest Territories, Canada
No full textClimate drivers of permafrost thaw are rapidly altering the stability of permafrost slopes across ice-rich terrains. The overarching goal of this thesis is to examine the setting, dynamics, and mechanisms of recent permafrost mass wasting in the central Mackenzie Valley of the Northwest Territories, Canada.
The first component of this thesis includes a literature review that emphasizes increasing occurrences of permafrost slope failures that have resulted in a greater variation in permafrost mass wasting. Critical geomorphic thresholds have been surpassed across contrasting ice-rich terrains, resulting in a geomorphic continuum of mass wasting processes. Variations in permafrost slope failure reflect thermal instability operating on the slope, where top-down (active layer detachment slides, retrogressive thaw slumps) and bottom-up (deep-seated permafrost landslides) mass movement styles produce distinct mass wasting forms, and an intermediary class where the thermal evolution of permafrost is more gradually modifying the behaviour of frozen slopes (frozen debris lobes). The range of variation in processes and morphologies, and the increasing complexity of mass wasting processes, can be broadly related to geological legacy, and controlled by ground ice, thermal, geomorphic, and ecosystem factors, and their interaction with climate drivers of permafrost thaw, such as temperature, precipitation, and surface disturbance.
The second paper from this thesis characterizes the regional drivers and spatial and temporal dynamics of permafrost mass wasting in the central Mackenzie Valley. Repeat satellite imagery and field observations were used to map the distribution and frequency of retrogressive thaw slumps and deep-seated permafrost landslides (n=276) that have occurred largely since 2004 in concert with increasing air temperatures and precipitation. Substantial increases in mass wasting frequency (278%) and magnitude (602%) have occurred from 2004 to 2020 and are comparable to mass wasting dynamics to other ice-rich landscapes of northwestern Canada. However, the abundance of deep-seated permafrost landslides is unique to the warm and thin permafrost of the central Mackenzie Valley and indicates a contrasting geomorphic response to permafrost thaw when compared to ice-rich regions where permafrost is cold and thick. Importantly, 82% of all failures mapped are within the limits of forest fire disturbances from the middle to late 1990s, indicating thermal preconditioning of landscape sensitivity to thaw.
The third paper examines the setting and failure mechanisms of two examples of deep-seated permafrost landslides recognized in the study area that do not conform to traditional understanding of top-down permafrost mass wasting. The geomorphology of these features and their geological and permafrost settings are characterized using field and remote sensing techniques, electrical resistivity tomography surveys, and thermal model simulations. These large (4.5 and 18 x 106 m3) and recent (failures in 2017 and 2018) landslides were found to be likely a result of climate warming, wetting and past wildfire disturbances contributing to increase ground temperature, alter slope properties, and cause bottom-up thaw, initiating movement of the entire permafrost layer. Subsequent tension cracks enable water infiltration to depth, driving feedbacks that accelerate bottom-up thaw leading to the rapid translation of the entire permafrost table into valley bottoms, leaving distinct deposits of frozen conical mounds (molards). This style of permafrost slope failure is termed a permafrost detachment slide (PDS) and note their importance as a distinct geomorphic response of thawing permafrost in warm (> −2°C) and thin (< 40m) permafrost settings.
The last paper focuses on a unique ground ice assemblage in ice-rich permafrost slopes, which involves repeated layers of slope-parallel ground ice interbedded with colluvium. This cryostratigraphic assemblage has been observed in the central Mackenzie Valley in association with deep-seated permafrost landslides, and rapidly retrogressing thaw slumps. Characterizing the type and setting of ground ice is a key control on the style and magnitude of thaw-driven permafrost mass wasting. The origin of this ground ice type from a site in the Klondike Region, Yukon, Canada, is inferred using a cryostratigraphic approach combined with computed tomography scanning, and stable O-H isotopic and dissolved ion analyses. A conceptual model explaining the ground ice genesis in these colluvial settings is developed that involves repeated instances of segregated ice formation separated by thaw-driven mass wasting on an aggrading permafrost slope