Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that often results in deficits in communication, social skills, and emotion regulation. Additional concerns include disruptions to the sleep wake cycle that results from circadian rhythm dysfunction. 40% of individuals with ASD also have clinically significant anxiety, which tends to exacerbate pre-existing behavioral issues and social deficits. Previous studies suggest an association between increased sleep dysfunction and increased issues with anxiety in typically developing (TD) adults, and have insinuated a possible bidirectional relationship between the two. This study aims to look at the relationship between sleep quality and anxiety in adults with and without ASD. 89 adults (ASD=39) from a longitudinal five-site NIH-funded study on sex differences in autism were included. Participants completed the Munich Chronotype Questionnaire (MCTQ), a measure of the amount of sleep, based on sleep-wake times, and the Screen for Adult/Child Anxiety Related Disorders for a measure of generalized anxiety. Analysis will include (1) independent sample t-tests to examine group differences in anxiety and sleep duration and (2) correlations between sleep duration (MCTQ) and generalized anxiety subscore from the SCAARED measure for the group with ASD and the typically developing group. I hypothesize that individuals with ASD compared to TD will demonstrate higher anxiety scores and worse sleep quality. I also hypothesize that there will be a correlation between higher anxiety scores and shorter sleep duration. Additionally we will explore sex differences in anxiety and sleep. If sleep quality is related to anxiety, this might support the increased use of sleep behavioral interventions to improve mental health in individuals with autism spectrum disorders.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that is characterized by difficulties in social communication skills, and repetitive restricted behaviors. ASD has been associated with alexithymia, or difficulties processing one’s and/or own emotions. Prior research suggests, individuals with alexithymia are at increased risk of developing disordered eating patterns, particularly if they are male (Shank, 2019; Larsen, 2006). Disordered eating tends to be associated with other psychopathology and possible to be comorbid with ASD. Individuals with ASD are also at increased risk of developing disordered eating (Huke, 2013), however it is unknown if the relationship between alexithymia and disordered eating is also found in autistic individuals. Thus, the aim of this study is to investigate sex differences in alexithymia and disordered eating patterns in individuals with and without ASD. 108 adults (ASD n=58, Male n=35) were included; all participants had an IQ ≥70. Alexithymia was measured using the Toronto Alexithymia Scale which is a 20-item instrument describing feelings, identifying feelings, and externally oriented thinking. Scores ≥61 indicate alexithymia . Eating behavior was measured using the Adult Behavior Eating Questionnaire which is a 35-item instrument involving 8 subscales including hunger, food responsiveness, emotional overeating, enjoyment of food, satiety responsiveness, emotional undereating, food fussiness, and slowness in eating. We expect to see a correlation between alexithymia and eating behavior in the autistic group, similar to what has been previously found in non-autistic groups. We also hypothesize that this correlation will be present for autistic males but not autistic females. This study will contribute to better understanding if alexithymia and eating behavior patterns are found in individuals with ASD, and whether or not sex mediates this relationship. If this relationship were true, this may affect how interventions for ASD are administered and label alexithymia as a risk factor for disordered eating.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by social communicative impairments and sensory sensitivities. Additionally, the physical and social changes that occur with puberty may be a turbulent time for adolescents. Earlier pubertal timing has been correlated with higher internalizing mental health symptoms for neurotypical girls with earlier onset of menarche in females being tied with higher rates of depression that persists into adulthood. This study investigates the relation between pubertal timing and internalizing mental health problems for autistic and non-autistic female adolescents in a longitudinal study. 23 female ASD participants (ages 8 to 17) and 42 female neurotypical participants (ages 8 to 17) from a NIH-funded project investigating sex and gender differences in individuals with autism are included. Participant data was collected at a second timepoint, 3 to 8 years later. Data on pubertal development was collected using the Pubertal Development Scale, a parent or self-report measure of physical development. Depression and anxiety were assessed using the Child Behavior Check List, a parent-report behavioral checklist of mental health symptoms at the first time point, and a self-report version of the CBCL at the second time point. First, we examine pubertal timing variation by calculating residuals of a pubertal maturation by time regression plot. Second, we will investigate the relationship between puberty timing and depression and anxiety using a correlation test. To further analyze this relationship between pubertal timing predicting future depression and anxiety, we will run a multiple regression test. I predict that the relationship between pubertal timing and depression and anxiety will be greater for autistic girls than for neurotypical girls. This study’s data can add a neurodiverse perspective on how pubertal timing impacts mental health in females and could provide evidence for the need of interventions and additional support to adolescent females with ASD.

Autism Spectrum Disorders (ASD) refer to neurodevelopmental difficulties in communication and social interaction. It is thought that 94% of autistic adults have used camouflaging behaviors at some point in their lives, meaning that they have developed certain behaviors to blend in the social world and to “hide” their autistic differences. Camouflaging behaviors include: masking - hiding the autistic features; compensation - practicing certain behaviors to compensate for certain social shortcomings; and assimilation - trying to fit in so they are not singled out (Hull et al., 2018). We are interested in the relationship between camouflaging behaviors and social communication in individuals with and without autism. Data from 85 participants (42 ASD, 48 females) ranging from 15 to 23 years old from the NIH funded study on sex and differences in autism were included in the analysis. Autism diagnosis was confirmed via standardized tests and all participants had an IQ of 70 or higher. Participants completed the Camouflaging Autistic Traits Questionnaire (CAT-Q), a 25-item questionnaire that tests the degree of using camouflaging strategies, and Vineland Adaptive Behavior Scales, a parental interview that informs the diagnosis of intellectual and developmental disabilities. We predict significantly higher camouflaging behaviors and lower socialization skills in the autistic group compared to the non-autistic group. We predict a positive correlation between CAT-Q scores and Vineland socialization scores in the autistic group, since by resembling their peers will make their parents report better social skills. We also predict that the correlation between masking and social skills will be higher in females than males in both groups, as females are found to have higher social motivations (Cook, Ogden, & Winstone, 2018). Camouflaging may prevent others from recognizing the symptoms of autism and fail to get diagnosis. Therefore, it is important to detect camouflaging behaviors so autistic children get timely treatments.

Current models of pain research involve restrictive forms of resident-intruder pairing where experimental mice are involuntarily placed in social situations. These methods have limited application as the research does not account for individual variability and the dynamic social decision-making characteristic of humans. Our research uses a novel volitional social procedure that more accurately represents human behavior in the context of pain. I conducted social self-administration protocols on C57 strain mus musculus to quantify changes in voluntary social interaction before and after neuropathic pain has been induced via spared nerve injury. In addition, I utilized a Von Frey filament test to measure changes in pain sensitivity over this time period. Two social self-administration (SA) experiments were conducted on separate cohorts of C57 mice. In Experiment I, SA was run intermittently at 3-day intervals following neuropathic injury, providing lapses between voluntary social engagement. In Experiment II, SA was run continuously following neuropathic injury. We found that the continuously run SA group experienced a rebound in social interaction to levels matching their pre-surgery states and sham controls, whereas the intermittent group displayed a stark decline in voluntary social interaction that reached statistical significance from sham controls on day 8. Interestingly, tests of allodynia that were conducted to determine prolonged mechanical sensitivity typical of chronic neuropathic pain showed that both groups were experiencing equal levels of increased pain sensitivity throughout behavioral testing. Our results show promise in revealing the dynamic connection between social interaction and pain perception. Research has already identified key areas of interest such as the medial prefrontal cortex (mPFC) and nucleus accumbens (NAc) as hubs responsible for regulating social behavior. We aim to further examine the physiological changes that occur in these areas as a result of persistent pain using a variety of sophisticated analytical techniques.

Activins are a class of growth factors belonging to the Transforming Growth Factor-Beta (TGF-β) superfamily that recruit a tetrameric complex of type-I (ACTRI) and type-II (ACTRII) transmembrane serine/threonine kinase receptors to cause downstream signaling in various critical cell-signaling pathways. Inhibition and supplementation of the signaling molecules offer new possibilities in many therapeutic and clinical applications. The overexpression of Activin A has been shown to be a major driver of diseases such as pulmonary hypertension, chronic kidney disease, and cerebrovascular disease. Sotatercept, one of the few existing activin-A-inhibiting therapeutics, is in clinical trials for treating pulmonary arterial hypertension, but, like many ligand traps, it is designed by fusing a large antibody domain to a native activin receptor domain, making it expensive to produce and unstable due to its large size. There is a need for therapeutics that can be produced in bacteria, replicate high efficacy and strong affinity, and minimize side effects. These therapeutics can be developed through de novo protein design, which involves designing proteins from scratch, with a new, unique sequence predicted to fold into its specific corresponding structure or by grafting native protein-protein interfaces onto novel scaffolds. This project aims to use computational de novo protein design methods to develop a therapeutic ligand trap for Activin A. The design pipeline involves scaffolding functional motifs, optimizing protein sequences, predicting protein structures, and filtering designs using metrics calculated by Rosetta, a protein design tool. With this pipeline, we have tested several design strategies, providing knowledge-based guidance and analyzing outputs to inform strategies for future design iterations. Recent designs have passed Rosetta filters, and we have ordered the corresponding genes to express and purify the designed proteins. Once purified, they can be experimentally characterized, testing binding and stability in vitro, and then further optimized to create stable Activin A ligand traps.

Autism spectrum disorder (ASD) is a developmental disorder characterized by deficits in social communication and interactions, and repetitive behaviors or interests. Children with ASD are more likely to show food selectivity than neurotypical (NT) children, and their restricted food consumption may be associated with nutritional deficiencies. Eating disorders are serious mental health conditions that can have long term health-related consequences. Understanding behavior surrounding eating habits is paramount to developing treatments that are effective for children with autism. In youth without autism, self-compassion impacts eating behaviors. Higher self-compassion is causally linked to lower eating pathology. This study aims to extend this work to examine whether self-compassion influences eating behaviors in adolescents with ASD. Participants included children with ASD (n = 37) and children with NT development (n = 52) between the ages of 13 - 17 years. Parents of participants completed the Child Eating Behavior Questionnaire, and participants completed the Self-Compassion Inventory for Youth survey. We compared emotional overeating and emotional under eating to the self-compassionate coping scale and self-punitive coping scale. Given the stereotyped and repetitive behaviors of adolescents with ASD, we expect a similar relationship between self-compassion and eating behavior seen in studies with neurotypical individuals, to also exist within an ASD group. Specifically, we hypothesize: (1) a positive correlation between emotional overeating and self-punitive coping, (2) a positive correlation between emotional under eating and self-punitive coping, (3) a negative correlation between emotional overeating and the self-compassionate coping scale, and (4) a negative correlation between emotional under eating and the self-compassionate coping scale. This study will offer more insight into the role of self-compassion and eating behavior of ASD adolescents. If the hypothesized correlations exist within the ASD and NT groups, we can look for similar risk factors of disordered eating behavior within ASD adolescents.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by differences in social interaction and communication, as well as repetitive behaviors (APA 2013). Per Kanne and Mazurek (2011), 56% of 1380 children and adolescents diagnosed with ASD engaged in some form of aggression towards family, teachers, or peers. Aggressive behaviors can prevent youth with ASD from being able to engage in learning opportunities and community events. An important factor in decreasing rates and severity of aggression is through the identification of social and environmental factors that may impact the stability of aggressive behaviors. The aim of this study is to look at factors that may impact the stability of aggressive behaviors in a longitudinal sample of autistic and non-autistic youth. 44 participants (22 ASD) aged 8 to 18 years, from the NIH funded longitudinal study on sex differences in autism were included in the analysis. Aggression was measured using the Child Behavior Checklist (CBCL), a parental report measure on problem behaviors, which includes items related to self aggression (self injurious behaviors) and other-aggression (aggression towards peers and adults). Data was collected at baseline and 3 to 8 years later. We expect low IQ and younger age will be related to higher aggressive scores on CBCL at baseline. We predict that youth with greater social improvement between timepoints and use of psychotropic medication will have lower levels of aggression over time. Identification of factors that impact changes in aggression over time can aid in implementing interventions to aggression and improve quality of life.

Voice assistants (VAs) are transforming how humans interact with technology. While promising, state-of-the-art VAs like Siri and Alexa do not incorporate a user’s spatiotemporal context such as their surrounding objects or gestures, which results in degraded performance and unnatural dialogue. Since pronoun usage is inherent to everyday speech, we expect future VAs to support ambiguous speech queries. We introduce GazePointAR, a wearable augmented reality (AR) system that resolves ambiguity in speech queries using eye gaze, pointing gestures, conversation history, real-time computer vision, and a large language model (OpenAI’s text-davinci-003). With GazePointAR, a user can ask “what’s over there?” or “how do I solve this math problem?” simply by looking and/or pointing. Upon voice activation, GazePointAR listens for the query, takes a screenshot, and narrows the focus by incorporating information from eye gaze, replaces the pronoun in the query with the detected objects and texts, and utilizes a language model to answer the modified query. To assist in this project, Liam and I reviewed relevant literature, brainstormed technical solutions for multimodal integration, constructed user study scenarios, and conducted reflexive thematic coding on qualitative data. To evaluate GazePointAR, we conducted a three-part lab study that compared GazePointAR to two other state-of-the-art query systems (Google Voice Assistant and Google Lens), examined GazePointAR’s pronoun disambiguation for three tasks, and concluded with an open-ended component where users could suggest and try their own queries. Participants appreciated the improved simplicity and human-likeness of context-aware queries; however, they preferred faster response times and better explanations for query results. By combining visual and voice inputs to answer a broader range of questions, GazePointAR provides a foundation for future works of VAs, such as designing a more anthropomorphic VA.

Cell division is essential to all living organisms, and microtubules are critical for this process. Microtubules in bundled kinetochore-fibers generate forces to move chromosomes by stochastically switching between assembly and disassembly states. Electron tomographs of mammalian mitotic spindles show that microtubule tips maintain a high level of coordination despite stochastic switching between states. The collective behaviors of microtubule bundles give rise to chromosome oscillation, which promotes the alignment and separation of chromosomes along the spindle equator during metaphase. To understand how stochastically-switching microtubules can produce synchronized motions in vivo, we use a mechanical model that describes the force-dependent dynamics of microtubules. The model reproduces the bistability of microtubule bundles observed in vivo and suggests that this coordination could be explained by sufficiently stiff mechanical coupling. We also propose a new way to characterize microtubule bundle state based on the substates of microtubules within the bundle. We will use this characterization to calculate microtubule bundle switch rates, which are technically difficult to distinguish otherwise, thereby laying the groundwork for future comparisons between bundle switch rates in vivo and in silico. In the future, we plan to explore the assumptions on which this model is based by analyzing the relationship between microtubule velocity and forces. We predict that this result will validate the assumed force-velocity relation of microtubule growth and contribute to a more accurate understanding of kinetochore-fiber behaviors. 

The Hijra is a South Asian community of genderqueer peoples which perform spiritual rituals and give blessings, typically at births and weddings in return for donations. This ancient community, however, has a long history of invisibilization and criminalization in South Asia. I am studying the Hijras’ physical and literary portrayal as it connects with, and unfolds from, their histories of criminality, invisibility, and mobility under imperial and post-imperial rule. Knowing this history is valuable to the field of colonial and gender/queer studies, as it provides a better understanding of how minoritized peoples navigate – and survive – changes in power and status under shifting power. The Hijra is a good interdisciplinary model for this, as they are minoritized at intersections of gender, religion, sexuality, caste, and class. I am using Indian epics, including the Mahabharata and Ramanayana, to theorize the figure of the Hijra as it exists in the dominant South Asian literary tradition. I am pairing this with articles on Hijra identity, which historicize the community’s legal and social statuses under the Mughal and British imperial formations, up into the decades post-South Asian independence. The purpose of tracing this history is two-fold: (1) to attempt the respectful re-tracing of a genealogy which, until now, has been invisible under various dominant empires and (2) to understand the Hijra peoples’ increasing inclusion into both Western research and South Asian democracy. The urgency and relevance of my inquiry has increased as pro-genderqueer legislation is currently being passed at unprecedented rates in South Asia. This includes India’s 2014 “third gender” law, which officially recognizes the Hijra as the nation’s “third gender.” While many celebrate this inclusionary shift, I argue that the change must be critically assessed, as it represents another incoming shift in global power that minoritized communities will have to navigate and survive.  

Refugee resettlement job training programs do not necessarily offer positive long-term economic outcomes. I have discovered a gap in peer-reviewed articles about the flaws within the existing resettlement employment services. The problem is that there are two types of employable refugees with distinct needs: (1) Refugees who already have professional credentials who need their credentials officially recognized so they may continue their chosen profession and (2) Refugees who need professional development that will connect them to sustainable paid employment. I have performed a systematic literature review that focuses on factors that contribute to the economic stability of the adult refugee population during their resettlement period in the United States. The selection criteria materials have been retrieved from the University of Washington Library Search and Web of Science including search terms related to refugee resettlement services that focus on the refugee population in the United States with referenced English language peer-reviewed scholarly articles published within the last five years. The overall findings show that refugees are to acquire basic job skills within the first six months of their resettlement as they are provided funding and housing during this short window of time. My research reveals that established programs only offer short-term solutions thus the significance of this problem is that the current system focuses refugees on survival employment with low paying jobs that do not necessarily match their qualifications. Refugees who are highly skilled and overqualified face insurmountable obstacles for credential recognition in the United States and are grossly underemployed, unable to sustain a livable wage. The implications of this have prevented both sets of refugees from reaching sustainable economic equity and have generational impacts on the livelihood of refugees in the country.

Management of muscle-invasive urothelial carcinoma is extremely limited with Anti-FGFR3 (Erdafitinib) being the only kinase-targeting therapy FDA-approved for treating advanced bladder cancer patients. Kinases represent an important family of proteins for drug development due to their well-characterized role in tumor growth and disease progression. Until the last decade, the process of identifying essential proteins was carried out by individually depleting their production using siRNA or shRNA knockdown. To expedite the process of discovery, we’ve applied a machine-learning model that predicts the activity of 428 kinase inhibitors and identifies the most essential kinases associated with promoting viability for five bladder cancer cell lines used in this study (COCAB1, SW780, COCAB11, UMUC5, & SCABER). This technique, known as polypharmacology, leverages the use of non-specific kinase inhibitors that target multiple individual kinases. We have validated the results of this computational method by measuring the effect on cell viability for 7 kinase inhibitors representing strong, moderate, and weak predicted impacts. Using live-cell imaging we quantified relative cell growth over 72 hours and found a positive correlation between predicted and observed effects on viability for all cell lines. The regression identified PTK5 (FRK) and VEGFR2 (KDR) as a common top essential kinase across all five cell lines. Currently, we are exploring validating the essentiality of these kinases by siRNA knockdown. Complete validation of this polypharmacology method would suggest continuing the evaluation of PTK5 and VEGFR2 as candidate novel therapeutic targets to treat advanced bladder cancer patients.

Many standard oxidants and reductants are non-reusable and toxic, so it is important to pursue cleaner alternatives. In this project, we have synthesized and characterized two metal-bipyridyl supramolecular cages and have studied their application as catalysts for the electrochemical reduction of organic substrates. Supramolecular cages are formed from the self-assembly of organic ligands and metal ions in solution, and they contain internal cavities with unique electronic microenvironments, similar to the interior of enzymes. While these polyhedral structures have been investigated as catalysts for traditional synthetic pathways, their role in electrosynthesis remains underexplored. Electrosynthesis involves the transfer of electrons to and from substrates using an applied potential, rather than chemical redox agents. This method is often hindered by a high kinetic barrier at the electrode-substrate interface, but catalysts can lower this barrier. We hypothesize that redox-active supramolecular cages – cages that can readily interconvert between charge states – can serve as effective electrocatalysts by encapsulating and transferring charge to substrates. To understand the effect of ligand geometry on electrocatalysis, I have synthesized two redox-active ligands with bipyridyl chelating groups. One contains a highly conjugated perylene core, while the other contains a compact core formed from pyromellitic dianhydride. We have metalated these ligands with iron ions to form two tetrahedral supramolecular cages. We then utilized cyclic voltammetry to assess cage-facilitated charge transfer to vicinal dihalide substrates. We observed that the reduction of multiple substrates, including 1,2-dibromo-1,2-diphenylethane, occurred at milder voltages in the presence of the cages, indicating a reduced kinetic barrier. For these substrates, we then performed bulk electrolysis, from which we determined that the percent conversion to the desired product was significantly higher when a cage was present, supporting our hypothesis. Ultimately, we aim to use these cages to enable electrosynthesis of organic feedstocks at lower voltages and with fewer byproducts.

Edentulism, or the loss of all natural teeth, is a condition typically resulting from dental caries and periodontal disease, described as the ultimate manifestation of oral health disease burden. Edentulism has significant effects on masticatory function, such as chewing, swallowing, and speaking, as well as mental health and overall quality of life. Food insecurity has been associated with adverse oral health outcomes. However, this relationship has been insufficiently studied among low-income older adults, who are disproportionately impacted by oral health disparities and more likely to be edentulous. We hypothesized that food insecurity would be associated with increased prevalence of edentulism among low-income older adults in Washington state. To test this hypothesis, we collected data at seven community-based sites across Washington, including the Multi-Service Center in Federal Way and the Pike Market Senior Center. We administered a survey containing a 10-item measure of food insecurity to 218 participants aged 50 years or older. We also conducted dental screenings that involved examination of the mouth, gums, and individual teeth. 45% of participants (n=98) had marginal, low, or very low food security. 7.8% of participants (n=17) had zero teeth. We utilized logistic regression models to examine the association between food security and edentulism. Our preliminary analysis indicates that older adults with low or very low food security had 2.44 times the odds of being edentulous compared to those with high food security, after adjusting for age, sex, and race. Further analysis can help inform the development of feasible and effective interventions to address oral health inequities among low-income older adults.

In humans, the KRAS protein normally acts as a switch to regulate cell growth. Acquisition of certain mutations result in constant activation of the protein, leading to uncontrollable cell growth. The most frequent such mutations are at the glycine-12 residue of the protein, including changes to valine (G12V). Cells with this mutation can be recognized and eliminated by T cells engineered to express an antigen-specific receptor (TCR), but posttranslational modifications, such as methylation, can interfere with the ability of such engineered T cells to recognize G12V mutant KRAS. The objective of my project is to identify and eventually disrupt mechanisms that lead to methylation of KRAS mutant antigens. I hypothesize specific enzymes are responsible for methylation of mutant KRAS. To accomplish this, I will generate CRISPR-Cas9 gene knockout libraries to screen for potentially responsible enzymes. The CRISPR library will target human genes encoding methyltransferase and demethylase enzymes. I will co-culture these CRISPR-Cas9 knockout cancer cells with T cells engineered with a TCR targeting the unmethylated KRAS mutant antigen. From the tumor cells that have survived this coculture, I will sequence genomic DNA to determine which knockouts are enriched/depleted after the co-culture. Preliminary results show certain gene knockouts are significantly enriched/depleted in co-culture compared to baseline. The sgRNAs that appear from the high-throughput knockout library most likely to be involved in the KRAS methylation pathway will be individually evaluated in additional co-culture experiments. This should allow me to assess the proliferation/susceptibility of these cancer cells in more detail. For confirmation of mechanism, I will evaluate the methylation status of the KRAS mutant antigen in cancer cells rendered susceptible. My goal is to determine the process leading to methylation of KRAS antigen and then target it to allow for more effective targeting of KRAS-driven cancers using TCR-T cell immunotherapy.

Microchimerism (Mc) occurs when a small amount of genetically different cells (or DNA) is acquired from another individual. Mc is acquired naturally during pregnancy due to exchange between the mother and child and can be detected decades later. Maternal microchimerism (MMc) is when a person harbors Mc from their mother. MMc is frequently detected in healthy adults but is increased in individuals with some autoimmune diseases, including scleroderma. Few studies have investigated MMc, especially whether it changes in a woman after her own pregnancies. One study tested MMc in peripheral blood of women during the time they were pregnant and occasionally detected MMc, but not if the woman had preeclampsia. No study has addressed whether MMc prevalence and quantities change in healthy women according to the time since the woman’s own childbirths or number of childbirths. This study addresses this knowledge gap. MMc was assayed using a panel of polymorphism-specific real-time polymerase chain reaction (qPCR) assays on DNA from peripheral blood. Most assays employed human leukocyte antigens (HLA)-specific primers and fluorogenic probes targeting non-inherited, non-shared HLA sequences. Each woman and her mother were HLA typed to identify an appropriate target. In total, 142 women were tested, and 266 qPCR experiments were run. Preliminary analysis found evidence of MMc in 59 of the 266 samples and a trend of MMc prevalence being highest within the first year postpartum and ten years after childbirth. Prevalence and quantities of MMc are being analyzed in collaboration with a biostatistician. Pregnancy and childbirth are known to affect some autoimmune diseases and cancer risk. Addressing the knowledge gap about MMc according to the time since birth and the number of births in women could provide further insights about some autoimmune diseases and cancers.

Hemagglutinin (HA) is a glycoprotein found on the surface of the influenza virus. HA binds to sialic acid receptors on host cells and mediates membrane fusion, allowing the virus to enter the cell. This project investigates how mutations associated with species crossover affect viral fusion mechanisms in influenza HA. HA from the past outbreaks of H5N1 influenza strains in Vietnam in 2004 (VN04) and Indonesia in 2005 (IN05), are being compared to an HA from an on-going avian influenza outbreak using a strain isolated in Colorado in 2022 (CO22). Through site-directed mutagenesis, mutations that affect acid stability and affinity for the human receptor were added to HA based on the VN04 and IN05 strains; we are studying these effects in recombinant protein rather than on infectious virus. These mutations are believed to enable the virus to increase transmissibility among mammals including humans. To compare how HA from these H5 isolates with and without the adaptive mutations behave, hydrogen-deuterium exchange mass spectrometry (HDX-MS) is being used to measure changes in deuterium incorporation on the protein backbone for specific peptide segments, giving a profile of local dynamics and structure throughout the HAs. By comparing the dynamic profiles for each as pH is lowered, mimicking acid-activation in host cell endosomes, we can probe how their structure in important fusion and human receptor-binding regions change as the fusion protein becomes activated. By comparing the structural dynamic changes of the WT and mammal-adapted, mutated IN05 and VN04 HA to the new CO22, we will be able to increase our understanding of the effect of the mutations that are associated with species crossover and hopefully be able to gain insight into the potential of this new avian influenza strain’s ability to become transmissible among humans.

Over the past 20 years, the capital city of Indonesia has sunk four meters into the ocean and continues at a rate of approximately 11 inches per year. At this rate, more than 95% of Jakarta, the capital city, is predicted to be completely submerged by 2050, inevitably drowning the city if stricter policies are not set in place for groundwater extraction. Due to the population’s poor access to piped water, private businesses and the public have resorted to illegal groundwater extraction, which has caused extensive land subsidence. The combined effects of regulated and unregulated groundwater extraction and the lack of equitable water access to the population have worsened the situation. Jakarta is not alone; other countries like China, Thailand, Vietnam, and the United States are also experiencing land subsidence due to extensive groundwater extraction. My literature review will demonstrate the effects of groundwater extraction and land subsidence in Indonesia by using a comparative public policy analysis on the different policies and management approaches of countries in Southeast Asia and North America. I will be comparing the policies in use by these different countries to determine the necessary policy reforms to create a more effective approach to solving the groundwater issue in Indonesia. I establish that the Indonesian government’s short-term plans to overcome land subsidence is not a sustainable solution to the damage that has been done to the city of Jakarta. Rather, the policies set in place need to be reformed to ensure that equitable water is supplied throughout the population and groundwater extraction is done safely and in moderation. This work is important not only to prevent the further sinking of Jakarta but also to overcome the long-term environmental impacts of groundwater extraction in Indonesia and other countries.

In long term data collection, sampling is a key parameter that can dictate the amount of data collected and also influence the available conclusions to be drawn. In the context of passive acoustic monitoring, recording an area’s soundscape at intervals (i.e., subsampling based on duty cycles) can alleviate data management costs, and has been widely investigated in the bird monitoring literature. However, the influence of duty cycle-based subsampling in passive acoustic monitoring of bats has not been thoroughly studied. Here, we discuss the effects of subsampling on ultrasonic recordings collected using AudioMoth recorders in the Union Bay Natural Area in 2022. We recorded continuously over the summer and then computationally applied duty cycles onto our data to generate subsampled data. The subsampled data was then fed into multiple bat call detection algorithms to understand the influence of different subsampling schemes. Our results show that subsampling schemes can impact the trends and activities that can be acoustically observed from echolocating bats. The results also show how the subsampling parameters may be tuned to collect valuable information while keeping data management costs low. We anticipate that this detailed investigation will aid in the design of efficient, long-term bat acoustic monitoring projects.

Genetically encoded fluorescent indicators (GEFI) change fluorescence level under microscope following a conformational change when bound to a target molecule, and can be used to visualize spatiotemporally specific biological processes involving a targeted molecule. Various imaging analysis tools exist to analyze the non-temporal fluorescent cell data, however there was no industry standard for the pipeline used to analyze molecular dynamics when imaged with GEFI in time-series experiments. This project aimed to develop a computational pipeline that analyzed the fluorescent readout of single cells in spatiotemporal experiments that utilized GEFI. The pipeline included both segmentation of cells, utilizing Cellpose, an existing deep learning-based generalizable and highly efficient segmentation program, and tracking of single cells across all frames. I personally contributed to the design, implementation, and testing of the tracking component of the pipeline. The tracking algorithm was designed using unsupervised machine learning, specifically k-means clustering with convolutional neural network feature extraction techniques. The pipeline was implemented using Python and made available and open source, accessible through Google Colaboratory for a more user friendly version, as well as Github for more thorough documentation and generalizability. Ultimately, this project aimed to minimize bias to result in more accurate and efficient high-throughput investigation of molecular dynamics when using fluorescent probes for dynamic cell imaging. Preliminary results demonstrated the effectiveness of the pipeline in tracking cells across various time points and provided a foundation for future optimizations and applications.