Texas Digital Library
UT Southwestern Medical Center Institutional Repository (University of Texas)Not a member yet
10274 research outputs found
Sort by
Deep Learning-Based Comprehensive Pathology Image Analysis
Get PDFThe advances in deep learning during the past decade have provided great tools for the analysis of histopathology images. Deep learning algorithms can aid in the routine diagnostics and have the potential to extract hidden information directly from slide images, providing valuable information for healthcare professionals.
This dissertation applied deep learning on the diseases of rhabdomyosarcoma and oral potentially malignant disorders. Convolutional neural network models were developed, one of which was to predict the risk of oral cancer development from slide images in patient with oral leukoplakia (a type of oral potentially malignant disorders), another to classify rhabdomyosarcoma histology subtypes and predict survival outcomes of embryonal rhabdomyosarcoma patients. Furthermore, U-Net and Mask R-CNN based HD-Staining models were used to identify cell nuclei of different tissue layers in oral epithelium, and novel Onion Peeling algorithm was developed to count the cell layer numbers.
Overall, the deep learning tools in this dissertation serve as aids to pathology evaluation process and can provide valuable information for risk stratification of patients
Kinetic and Thermodynamic Allostery in the Ras Protein Family
No full textThe general metadata -- e.g., title, author, abstract, subject headings, etc. -- is publicly available, but access to the submitted files is restricted to UT Southwestern campus access and/or authorized UT Southwestern users.An understanding of allosteric communication is necessary for interpreting protein function and regulation. For proteins involved in disease, such knowledge can lead to discovery of new sites for therapeutic targeting. However allosteric mechanisms have proven to be diverse, and allosteric communication in many proteins cannot currently be explained by their structure or dynamics. Progress has been made in elucidating the ensemble nature of allosteric communication, especially using MD simulations to provide structural specificity of previously averaged conformations. Here I show how kinetic (i.e. temporal) correlations encode information that is orthogonal to existing work on thermodynamic correlations. I performed atomistic simulations on H, K, and NRas isoforms in various states in the Ras cycle, for a total of 0.5 milliseconds. I show that Ras' most important structural motifs, switch I and switch II, are the primary members of my calculated thermodynamic and kinetic allosteric networks, consistent with the known roles of these two motifs in Ras function. I also reveal how these communication networks are altered by the presence of the -phosphate, as well as binding of the downstream effector Raf kinase. Strikingly, these communication networks correspond to structural motifs that are functionally engaged in the Ras cycle step being simulated. I find that kinetics-based allosteric communication is not restricted to the boundaries of secondary structure elements and can occur across long distances. I show that known features of Ras regulation, such as activation of allostery upon Raf binding, necessitate kinetic allostery. These data explain experimentally observed allosteric relationships by revealing the kinetic and thermodynamic communication pathways, and show how both modes of communication are needed to relay information within proteins. This work suggests that kinetics-based communication is a key mechanism of protein function and regulation
Structural Studies of Early Eukaryotic Ribosome Biogenesis
Get PDFRibosomes are macromolecular machines that synthesize all the proteins in our cells during translation. These complex molecules consist of ribosomal RNA (rRNA) and ribosomal proteins (RPs), which undergo multiple tightly regulated rearrangements before they achieve a fully active, mature conformation. This process is called ribosome biogenesis. Eukaryotic ribosome biogenesis begins with the transcription of ribosomal DNA (rDNA) followed by multiple rRNA assembly events, extending from the nucleolus to the cytoplasm and involving more than 200 proteins. It is carried out in two distinct pathways: one for the large subunit (60S) and one for the small subunit (40S). Once they achieve their mature conformation in the cytoplasm, the subunits bind to each other and form a mature ribosome (80S). Since the linear rRNA needs to be quickly assembled into a compact tertiary structure before export from the nucleus, most of the rearrangements occur in the nucleolus and the nucleoplasm, involving numerous types of enzymes, such as DEAD-box helicases, ATPases, methyltransferases, GTPases and others. Because the function of these factors remains poorly understood, my goal was to investigate their role in ribosome biogenesis by focusing on their enzymatic activities. Here, I present structural and genetic data demonstrating the functional relationships between different classes of enzymes, which are essential during early maturation steps. Firstly, I show the final nucleolar assembly events prior to the transition to the nucleoplasm, including a detailed mechanism by which Spb4, a DEAD-box protein, recognizes and unwinds its rRNA substrate, which in turn prepares the binding site for the Spb1 methyltransferase domain (Spb1-MTD). Spb1 methylates G2922 in the rRNA helix 92 (h92) and our cryo-EM studies reveal that Spb1-MTD is stabilized by a structural protein- Rrp17. Finally, I demonstrate that Spb1 enzymatic activity regulates the function of a downstream GTPase Nog2 in the nucleoplasm, and by using cryo-EM, I determine the mechanism of Nog2 GTP hydrolysis in a G2922 methylation deficient background
Dust
Get PDFThe author submitted this entry in the 10-Word Story category (Amateur division) for the 2025 On My Own Time™ (OMOT) Art Show.Inspired by the fleeting nature of memory
Establishing Neurocognitive Profiles of Major Depressive Disorder Subtypes
Get PDFMajor Depressive Disorder (MDD) subtypes have been shown to differentially impact psychiatric symptom presentation, clinical features, and functional abilities. While there is extensive research regarding MDD subtypes and clinical characteristics, there has been limited information regarding the relationship between MDD subtypes and neurocognitive functioning. In particular, the neurocognitive impact of the treatment resistant depression (TRD), defined as MDD that is unresponsive to treatment, and psychotic depression, defined as a depressive episode with hallucinations and/or delusions, is unknown despite the significant illness burden these subtypes represent. The aim of this dissertation was to address this gap by characterizing the neurocognitive profiles of TRD and psychotic depression across multiple neurocognitive domains and compare the magnitude of impairment between the two.
Data utilized were drawn from a broader NIMH-funded, randomized, controlled study conducted at the University of New Mexico that investigated the clinical and cognitive outcomes of varying pulse amplitudes during acute electroconvulsive therapy (ECT) in adults with MDD. Participants in the study were age 50+ with a diagnosis of MDD, and further delineated by subtype as TRD and psychotic depression. For this analysis, we utilized demographic and baseline neurocognitive data collected prior to start of treatment. Neurocognitive measures included the Montreal Cognitive Assessment (MoCA), Delis Kaplan Executive Function System (D-KEFS) Verbal Fluency and Color-Word Interference Subtests, Hopkins Verbal Learning Test-Revised (HVLT-R), and the Wechsler Adult Intelligence Scale 4th Edition (WAIS-IV) Digit Spans. Results reflected distinct neurocognitive profiles for the TRD and psychotic subtypes, with the psychotic depression group demonstrating a greater magnitude of impairment across neurocognitive domains. These findings suggest further research is warranted regarding the neurocognitive impact of MDD subtypes
Adaptive Behavior in a Sample of Culturally and Linguistically Diverse Children with 22q11.2 Deletion Syndrome
Get PDF22q11.2 Deletion Syndrome (22q11DS) is one of the most common genetic causes of developmental delays. Early identification and diagnosis of 22q11DS are essential to facilitate prompt therapeutic interventions; however, inequities in healthcare such as access to specialists and genetic testing often delay diagnosis and therapies. The diverse clinical manifestations of this multi-systemic disorder may involve medical conditions, neurodevelopmental and psychiatric disorders, known to impact quality of life, wellbeing, and functional outcomes of people living with 22q11DS and their families. The first aim of this dissertation (Study 1) was to conduct a scoping review on adaptive functioning across the lifespan in people with 22q11DS, compiling information about factors used to contextualize the participants' functioning. Findings revealed mixed results in the literature, largely focused on intellectual/emotional functioning and their impact on adaptive functioning. Common associations were found between impaired adaptive functioning, low intellectual functioning, and negative mental health symptoms (e.g., social withdrawal). However, studies relied on clinical populations and used different methodologies (e.g., measurements). Moreover, research about contextual/psychosocial factors or functional outcomes is scarce. The second objective of this dissertation (Study 2) was to use archival clinical questionnaires and medical records to explore individual, familiar, and societal level factors associated with adaptive behavior in racial/ethnically diverse patients with 22q11DS (n= 51) followed at a multidisciplinary clinic in Texas. Participants averaged low-to-below average scores on measures of adaptive behavior, with no significant associations with sex or externalizing/internalizing emotional problems. Low intellectual functioning, behavioral problems (e.g., withdrawal), and family functioning, were significantly associated with lower adaptive behavior scores, particularly in low-income families. Although limited by a small sample size and caregiver-reported data, the study underscored the need to monitor early changes in behavior and family functioning. Future studies should investigate potential mechanisms through which access to resources could be protective for adaptive behavior
[NOON MEAL.]
Get PDFThe author submitted this entry in the Fictional Short Story category (Amateur division) for the 2025 On My Own Time™ (OMOT) Art Show.An emotional patient with a heart-wrenching story. Next to her, a lunch tray and the receipt detailing her [NOON MEAL.
My Mind
Get PDFThe author submitted this entry in the Open Verse Poetry category (Amateur division) for the 2025 On My Own Time™ (OMOT) Art Show.This poem is a commentary on what it is like to live a life for the comfort of other people as a neurodivergent person. There are chronic, daily adjustments for the those around you who might otherwise judge your behaviors without having compassion or curiosity into why you might behave, need, feel, or communicate different than most. This poem is about years of building yourself based on criticism, and still not getting it "right". But you strive to make those changes anyway because at the end of the day, just like any other human you just want to be feel connected and known. This poem was written without the letter "e", the most common letter in the English language, as part of a series on identity
From Song to Sleep: A Multi-Project Investigation into the Courtship Song of Zebra Finch, Memory Consolidation, and Transgene Expression
Get PDFAt the core of neuroscience lies the fundamental question of how neural circuits integrate and execute behaviors. Decades of research have been devoted to understanding the neuroethology of vocal imitation. As such, my doctoral dissertation centers around broadening the frontiers of research by employing innovative and comprehensive methodologies. Here I describe a hybrid machine and deep learning pipeline I have developed to investigate song behavior at scale. I then share my findings regarding the identification of key components of song important for mate selection. Additionally, I report on novel viral constructs that have been designed to study the fundamental neurobiology underlying vocal learning and memory. Last, I present my finding on a neural correlate of sleep-based memory consolidation
Personalized approaches to obesity treatment
Get PDFDetailed formal protocol with illustrations and extensive bibliography.A recording of the protocol presentation is available on UT Southwestern’s Mediasite. Note: Access to the video is restricted to authorized UT Southwestern users only.UT Southwestern--Internal Medicin