136879 research outputs found
Sort by
John Bickham field notebook: AK2501-AK3000.pdf
No full textBound book, each page corresponds to a karyotype slide data.Data pages for AK3001-AK3500 corresponding to unique identifiers of specimens/samples examined for biological research. Specimens are primarily housed at Texas A&M University; Biodiverstiy Research and Teaching Collection
Quantifying Complex Grazing Management Practices and Producer Systems Thinking Skills
No full textThe capacity of alternative grazing management strategies to maintain or enhance the production of ecosystem services and ranch profitability continues to be vigorously debated. However, the parameters used to differentiate among alternative grazing systems are inadequate, and standardized methods for comprehensively characterizing grazing systems are lacking. Inconsistent approaches to, and definitions of, adaptive management have likely contributed to the inconsistent findings in grazing management research. Systems thinking has been promoted to enhance decision making in complex natural resource systems but there is a lack of research evaluating the relationship between adaptive management implementation and systems thinking skills. This dissertation (1) develops a rigorous weighted composite index to serve as a standardized approach for more accurately classifying the grazing intensity implemented in grazing management systems; (2) develops an instrument that thoroughly measures the implementation of adaptive management at the ranch level, and; (3) tests the hypothesis that the level of adaptive management is positively associated with a producer���s level of systems thinking skills. The Grazing Intensity Index (GII) characterizes grazing systems differently than characterization efforts that rely on only a few descriptors, which suggests that the GII more objectively and comprehensively captures the cumulative effect of the spatiotemporal distribution of grazing and rest within a grazing system. The Adaptive Management Index (AMI) is positively correlated with perceptions of increased connectedness between grazing system elements, which suggests that the AMI effectively measures the implementation of adaptive management in livestock grazing systems. Gross livestock sales and gender, rather than the AMI, are primary drivers of systems thinking as measured by the systems thinking skills instrument, suggesting that producers who generate a larger volume of sales from their operation have a greater tendency for viewing systems holistically. Implementing the GII and AMI in future grazing management research studies will enable researchers to more precisely characterize contrasting grazing management strategies resulting in more robust findings and enhanced communication among researchers, and between researchers, extension personnel, and producers. Additional research is needed to evaluate the relationship between adaptive management implementation and systems thinking skills
Geometric Deep Learning for Science: Prediction, Generation, and Symmetries
No full textDeep learning has significant potentials in accelerating the progress of science research. However, the data in most science problems are geometric data, or graph data, which brings many unique challenges. First, designing label-invariant data augmentations for geometric data is challenging. Second, regular deep generative models need to be dramatically modified to suit for 2D molecular graphs, 3D molecular geometries, and periodic materials. In this dissertation, we study these challenges and propose several novel methods to tackle them.
We first propose GraphAug, a novel automated data augmentation method aiming at computing label-invariant augmentations for graph classification. GraphAug uses an automated augmentation model to avoid compromising critical label-related information of the graph, thereby producing label-invariant augmentations at most times. To ensure label-invariance, we develop a training method based on reinforcement learning to maximize an estimated label-invariance probability. Second, we propose GraphDF, a novel discrete latent variable model for 2D molecular graph generation based on normalizing flow methods. GraphDF uses invertible modulo shift transforms to map discrete latent variables to graph nodes and edges. We show that the use of discrete latent variables reduces computational costs and eliminates the negative effect of dequantization.
Third, we propose G-SphereNet, a novel autoregressive flow model for generating 3D molecular geometries. G-SphereNet employs a flexible sequential generation scheme by placing atoms in 3D space step-by-step. We propose to determine 3D positions of atoms by generating distances, angles and torsion angles, thereby ensuring both invariance and equivariance. In addition, we propose to use spherical message passing and attention mechanism for conditional information extraction.
Finally, we propose SyMat, a novel symmetry-aware periodic material generation method. SyMat generates atom types and lattices with a variational auto-encoder model. In addition, SyMat employs a score-based diffusion model to generate atom coordinates based on a novel coordinate diffusion process. We show that SyMat is theoretically invariant to all symmetry transformations of materials.
We demonstrate the effectiveness of our proposed methods with comprehensive benchmark experiments. In the future, we will explore developing novel predictive models for the prediction of Hamiltonian matrices and accelerating the generation of SyMat by stochastic differential equation based diffusion models
Building 0D and 2D Porous Metal-Organic Nanomaterials for Efficient Photo-Induced Energy Transfer
No full textThe concept of host-guest chemistry has been raised to study the interaction between a system and attached small molecules. These strong interactions have enabled the use of framework materials in catalytic reactions with high selectivity and turnover numbers. Recently, the introduction of metal/metal clusters has revitalized this field due to improved stability and binding capabilities. Metal-organic frameworks (MOFs), renowned for their high crystallinity, porosity, and well-determined structures, have been extensively used for host-guest studies. However, the traditional 3-dimensional bulk MOFs create a diffusion barrier that hinders the applications. To overcome this, chemists have formulated the field of 0-dimensional molecular cages, termed ���0D porous coordination cages���, which maintain homogeneity and exhibit explicit confinement in all dimensions, and the field of 2-dimensional MOF-derived nanosheets, termed ���2D MOF nanosheets���. These cages facilitate efficient catalysis by avoiding 3D stacking of pores.
In my PhD research, I aim to develop viable synthetic methodology for 0D and 2D metalorganic nanomaterials. Embedding photoactive ligands into 0D cages and 2D MOF nanosheets provides an approach to introducing catalytic ability. This can be achieved through introducing coordination functional groups such as carboxylate and azolate groups, converting various photoactive molecules into suitable ligands while maintaining their activities. Moreover, the selection of coordination/metal clusters and the host-guest interaction in these materials can lead to distinct activities. This work launched studies on the fabrication of 0D cages and 2D MOF nanosheets and their photo-catalytic reactions. The objective of this research is to study the design rationales of 0D and 2D MOF nanomaterials and inspire the discovery of novel catalysts with high selectivity and activity
A Study of Texture Characterization of Fiducial Markers for Visual Navigation
No full textA full six degree-of-freedom state estimation is an important problem in robotics, augmented reality, and autonomous navigation. Obtaining such information using visual features has been challenging since such a task always needs information-rich images. Getting access to a visually rich environment is not often possible. In such scenarios, artificial visual references like fiducial markers are used. This thesis conducts a systematic study of the texture of such fiducial markers. The study provides insight into the fiducial markers��� visual characteristics and design patterns. A general implementation of the ArUco marker detection and estimation system is created to understand the fiducial marker design process fully. The thesis reports results and lessons from both the set of tasks. Based on the study of other markers, a new fiducial marker-based on the Spidron pattern is proposed. A detection and pose estimation system is developed for this Spidron based marker. The system is tested in an experimental setup simulating refraction, motion blur due to rotation and jitters, and transformation due to scaling
Data Modeling, Computing, and Generation: New Techniques by and for AI
No full textThis dissertation investigates approaches in data handling within the domain of Artificial Intelligence (AI), covering data modeling, computing, and generation. It explores four primary tasks, each addressing distinct challenges and presenting novel solutions in their respective fields.
In the realm of data modeling, the Side Information Boosted Symbolic Regression (SIBSR) and Symbolic Modeling techniques are introduced. SIBSR incorporates side information into the symbolic regression process, enhancing the search for accurate mathematical relationships in complex datasets. Symbolic Modeling extends this approach to multi-dataset scenarios, particularly in financial asset pricing, providing adaptable and interpretable models that capture the dynamics of financial markets.
For data computing, the focus shifts to neuromorphic systems with the analysis of new Analog Error-Correcting Codes (ECCs) and the design of neural network-based decoders. These advancements address the challenges of reliability and accuracy in analog data processing, marking a progression of error-correcting from digital to analog and benefits in neuromorphic computing environments.
In data generation, Reinforcement Prompting, a novel methodology that leverages Large Language Models (LLMs) for the generation of synthetic data, is proposed. This approach mitigates issues of data privacy and scarcity of labeled datasets, especially in the finance domain. This method demonstrates that models trained on the generated synthetic data maintain performance integrity comparable to those trained on real financial data.
The dissertation presents a comprehensive exploration of these methods, substantiated by experimental evaluations and theoretical analysis. The research contributes to the advancement of AI in data handling, offering new perspectives and tools in data modeling, computing, and generation. The findings underscore the transformative potential of AI in understanding, processing, and generating data more effectively and ethically across various domains
A Tide That Raises All the Boats: The Soviet Threat, NATO, and Marine Corps Innovation, 1969-1991
No full textAfter suffering the bitter defeat of the Vietnam War, the United States Marine Corps entered a nearly two-decade long period of reform and modernization. By the end of the Cold War in 1991, the service had markedly improved its training, equipment, doctrine, and proficiency for mid- and high-intensity conventional conflicts. Notably, the capabilities gained during the 1970s and 1980s remained in place well into the twenty-first century, marking this era as one comparable to the service���s most storied periods of innovation. The existing historical literature of the late-Cold War Marine Corps ascribed the service���s rise either to its military culture or crises in Southwest Asia. Such interpretations, however, overlook the impetus provided by the Soviet military threat and North Atlantic Treaty Organization (NATO) missions. An analysis of multi-archival and multi-service sources ��� including declassified war plans and studies, strategies and concepts, congressional testimony, exercise reports, student papers, journal articles, memoirs, oral histories, interviews with participants, commentary from NATO allies, and related secondary histories ��� demonstrates that the Soviet threat and NATO missions were salient driving influences on Marine Corps innovation during this period. This impetus, ever present, often unacknowledged, and occasionally even strenuously denied, provided the service with a unique azimuth of innovation, one qualitatively different from the id��e fixes of other militaries. The resulting service strategy served as a ���tide that raises all the boats,��� a focus on the severest test ��� a Soviet anti-amphibious defense in Europe ��� that simultaneously rehabilitated the Corps��� political relevance, modernized readiness, and inspired future concepts, all while still preserving flexibility for global employment across the spectrum of conflict. Accordingly, at Cold War���s end, observers praised the Corps as the ���most general-purpose force of the general-purpose forces,��� a service particularly well-suited for the uncertain new security environment. This dissertation advances historical interpretations of the Marine Corps and military innovation in the late-Cold War and demonstrates the effectiveness of threat-based strategies of innovation, providing a case study of value both to scholars of military innovation studies and strategic practitioners
Genetic Ablation of PITPNC1 Impairs Proliferation and Tumor Growth in Murine Melanoma via Mitochondrial Dysregulation
No full textPhosphoinositides (PIPs), the phosphorylated derivatives of phosphoinositols (PtdIns) are critical for eukaryotic cellular signaling. Major dysregulation of PIP signaling pathways has catastrophic effects, as subtle changes can cause neurodegenerative diseases and cancer in mammals. In this context, studying phosphatidylinositol transfer proteins (PITPs) urges our interest. Recent studies have indicated an important role for PITPs in determining the biological outcomes of PIP signaling, via a novel mechanism in which they likely do not function as lipid carriers.
PITPNC1, a member of under-investigated mammalian phosphatidylinositol transfer proteins (PITPs), is overexpressed in several metastatic tumors, including breast, colon, pancreatic, and melanoma. The tumor suppressor micro-RNA miR-126 also identifies PITPNC1 as a crucial target, of disease relevance. However, to date, no studies have emphasized the role of PITPNC1 in a complete knockout and immunocompetent environment. This has left critical gaps in understanding the role of protein in the signaling and metabolism of cancer cells and addressing these issues remain our primary focus. Our research demonstrates that PITPNC1 knockout mice are born alive and do not display overt phenotypes. However, phenotypic characterization of PITPNC1-/- mouse reveals changes in body fat and non-shivering thermogenesis in the females. PITPNC1 is associated with reduced patient survivability in melanoma cancer and its expression correlates with metastatic progression in murine and human melanoma cells. To understand the role of PITPNC1 in the context of melanoma, we employed a syngeneic murine melanoma model in which B16 melanoma cells were introduced to immunocompetent mice by subcutaneous injection. We demonstrated that PITPNC1 overexpression promotes tumor growth and metastasis in melanoma tumors. In addition, CRISPR-Cas9 generated PITPNC1 knock-out melanoma cells inhibit melanoma cancer progression in vivo in both non-cell-autonomous and cell-autonomous manner. Next, our findings indicate that PITPNC1 deletion alters the lipidomic profile in B16F10 melanoma cancer with loss of PIP3 lipid and cholesterol esters and accumulation of acylcarnitine and triglycerides, suggesting compromised fatty acid oxidation, which potentially underlies mitochondrial dysregulation in the PITPNC1 null condition. Lastly, we demonstrated that PITPNC1 null cells have a defective mitochondrial function with increased mitochondrial fragmentation in B16F10 melanoma cells. In this study, we conclude that PITPNC1 overexpression promotes tumor metastasis while genetic ablation of PITPNC1 impairs tumor growth in murine melanoma via mitochondrial dysregulation
Dynamic Covalent Polymer Networks for Additive Manufacturing and Supersonic Impact
No full textThe use of dynamic covalent bonds enables self-healing, shape morphing, and energy dissipation in materials. During the past decade, significant progress has been made in advancing both network chemistry and structural design. Diels-Alder (DA) ���click��� reactions introduce room-temperature stability and thermal response to the polymer networks. This thesis explores the roles of temperature and stress in the unique behavior of DA dynamic polymer (DAP) materials for their use in additive manufacturing and supersonic impact. In the second chapter, we report highly conductive nanocomposites made of DAP networks containing branched multi-wall carbon nanotubes (b-CNTs). The ability of liquified DAP to wet, infiltrate, and chemically stabilize bCNTs at increased temperature, and then ���lock��� well-dispersed nanotubes upon cooling via the ���click��� DA reaction results in a low percolation threshold of 0.04 wt% b-CNTs. Moreover, the nanocomposites exhibit controlled network plasticity for permanent shape reconfiguration at solid state via Joule-heating-induced dynamic bond exchanges, and these transformations can be triggered selectively at different locations in printed multi-material hybrid constructs, enabling spatiotemporal control of the material's permanent shape. In the third chapter, we demonstrate that large puncture healing of ultra-thin DAP dynamic networks under supersonic impact by microprojectiles outperforms that of traditional glassy polymers, while showing energy absorption comparable to those materials. Post-mortem microscopies reveal efficient puncture healing that is likely enabled by stress- and temperature-induced viscoelastic responses of DAP networks. The progression of impact events was observed using in-situ imaging with a nanosecond, microscale resolution, while the recovery of DA bonds after the impact was confirmed by infrared nanospectroscopy. In the fourth chapter, we present a procedure design for manufacturing DAP microspheres and analyze their high-strain-rate deformation behaviors under supersonic impacts against a rigid substrate. The dynamic responses of these microspheres are associated with the viscoelastic and viscoplastic characteristics of the microspheres as well as interfacial adhesion. Unlike traditional polymers, the interfacial adhesion properties are governed by the thermomechanical responses of DAP networks, due to temperature- and stress-sensitive DA bonds. The role of thermomechanical network responses in microsphere impacts was unveiled by in-situ observations, post-mortem morphological analysis, and finite element analysis simulations
Probing Membrane Protein-Ligand Interactions Using Native Mass Spectrometry
No full textThe cell membrane, essential for cellular structure and function, is composed largely of membrane proteins and lipids. While transmembrane proteins are known for their critical roles in various physiological processes such as maintaining cell resting potential and facilitating molecule transport, the specific interactions between these proteins and different lipid moieties are not fully understood. In this study, we employed native mass spectrometry as a primary tool to unravel these interactions, capitalizing on its unique ability to preserve non-covalent interactions in biomacromolecules.
Our investigation comprised two distinct approaches: single ligand screening and multi-omic screening. Through single ligand screening, we identified specific lipid moieties that bind to the mammalian two-pore domain potassium channel TRAAK. This approach not only pinpointed the preferred lipids but also revealed a dose-response relationship with TRAAK potassium efflux in functional assays. On the other hand, our multi-omic analysis yielded significant findings. It corroborated the unique allosteric modulation observed between cardiolipin and phosphatidylethanolamine in their interaction with the E. coli ammonium channel AmtB, aligning with previous research. Additionally, this approach led to the discovery of a complex allosteric modulation involving TRAAK, phosphatidylserine, and cupric ions. Impressively, these interactions were maintained when ejected from proteoliposomes, bridging our functional assays, and validating our methodologies.
Moreover, our research broke new ground with the synthesis of novel charge-reducing molecules. This advancement has significantly enhanced the study of less stable membrane proteins, paving the way for more comprehensive exploration of membrane protein-lipid interactions. Overall, our findings contribute substantially to the understanding of cellular membrane dynamics and functionality, marking a significant stride in the field of cellular biology