Serotonin 1B receptors (5-HT1BRs) are involved in cocaine reward via regulating activity of dopamine neurons. The 5-HT1BR agonist CP-94,253 or 5-HT1BR overexpression in the nucleus accumbens shell (NAcSh) enhances cocaine intake during maintenance of daily self-administration (SA) but inhibits intake after 21 days of abstinence in male rats. My central hypothesis is that CP-94,253 acts at 5-HT1BRs located on the terminals of NAcSh GABA neurons that undergo regulatory changes in response to cocaine SA and subsequent abstinence resulting in an abstinence-induced switch in the functional effects of CP-94,253 in both male and female rats. In the first series of experiments, I compared the functional effects of CP-94,253 in female rats to male rats: 1) during maintenance of daily cocaine SA, 2) after 21-60 days abstinence, and 3) during the resumption of cocaine SA after abstinence (i.e. model of relapse). I found that CP-94,253 enhanced cocaine intake and breakpoints on a high-effort progressive ratio schedule of cocaine reinforcement during maintenance regardless of sex. By contrast, CP-94,253 attenuated cocaine intake after 21 days of abstinence and during the relapse test, regardless of sex. These findings suggest: 1) an abstinence-induced inhibitory effect of the 5-HT1BR agonist occurs in both sexes, 2) these inhibitory effects are long-lasting, and 3) the agonist may provide a novel therapeutic for cocaine use disorders. I next used RNAscope in situ hybridization to measure regulatory changes in 5-HT1BR mRNA expression and its co-expression with GABAergic and glutamatergic cell markers in the lateral and medial NAcSh subregions after abstinence from cocaine. I found no significant changes in these measures in either subregion of NAcSh after prolonged abstinence in either sex; however, I did observe that 95% of 5-HT1BR mRNA is co-localized in GABAergic neurons, whereas <2% is co-localized in glutamatergic cells. Future research investigating abstinence-induced, functional changes in 5-HT1BRs in subregions of the NAcSh is an alternate approach to further test my hypothesis. This research is important for the development of 5-HT1BR agonists as putative treatments of cocaine use disorders.

Pure metal clusters serve as model systems by providing an avenue for the study of fundamental phenomena, specifically the interaction between light and matter. Bulk metal materials are known to display defining characteristics, namely thermal conductivity, electrical conductivity, and luster, which provide a quantifiable measure of their metallicity. These properties are all due to the electron delocalization throughout the metal. Nanoscale materials lack the ability to measure these properties, leading to the need for a manner of quantifying the metallic character at the nanoscale size regime.Excited state lifetimes vary for semiconducting and metallic systems, specifically metals relax to a ground state at a faster rate than semiconducting materials. Aluminum clusters have received decades of attention regarding their metallicity. Moreover, Al clusters have been debated to fit into the jellium model. The jellium model seeks to describe a cluster as a “superatom” where all electrons are delocalized around the positively charged metal center, like that of an atom. With three valence electrons, jellium shell closings can be met if the electrons involved in cluster bonding varies. This variance leads to a localization of electrons for instances in which all three electrons do not contribute to bonding. Localized electrons aren’t characteristic of the jellium model or metals more broadly. Tracking the excited state lifetimes of Al clusters produced through laser ablation seeks to uncover the onset of metallic character. Femtosecond pump-probe spectroscopy coupled with time-of-flight mass spectrometry has resolved the time dynamics for atomically precise Al clusters ranging in size from 1-43 atoms. At a size greater than 9 atoms, it’s identified that Al clusters show metallic character. This finding is supported by previous literature results and the fact that, above 9 atoms, Al cluster excited state lifetimes match that of the bulk scale Al excited state lifetime of ~300 fs.

Peatlands are significant global carbon sinks, where plant litter accumulation outpaces the rate of microbial degradation, which can result in significant emissions of methane and carbon dioxide. The Pastaza-Marañón foreland basin (PMFB) in the western Amazon contains the largest expanse of tropical peatlands in South America, characterized by a diversity of soil properties, including pH and mineral concentration. The PMFB is predicted to decrease in its carbon capture capacity along with a rise in greenhouse gas emissions as the climate changes. Therefore, it is imperative to understand the impact that soil properties have on the abundance of functions, microbial physiology, and interspecies interactions between microbial community members. Metagenomic sequencing of soil samples from three geochemically distinct peatlands revealed site-specific enrichment of functions related to carbon, nitrogen, phosphorus, and sulfur cycling. Additionally, 519 metagenome-assembled genomes (MAGs) were recovered, revealing variations in microbial populations responsible for organic matter degradation and nutrient (nitrogen and sulfur) cycling across sites. From these MAGs, a novel family within the Bathyarchaeia was identified, Candidatus Paludivitaceae. This family is putatively capable of carboxydotrophy, able to use CO for energy and biomass. Subsequently they could detoxify the environment of CO benefiting other community members and playing an indirect role in modulating carbon cycling. To experimentally investigate interactions of peatland microbes, co-culture experiments assessed the impact of carbon substrates (4-hydroxybenzoic acid, mannitol, and arginine) on microbial interactions from heterotrophs isolated from two geochemically distinct peatlands. Results indicate substrate and peatland type significantly influence nature and frequency of microbial interactions. The response of microbial genera to carbon substrate also varied showing the role of metabolic traits and substrate preferences in determining growth patterns of microbes. This research advances our understanding of microbial ecology in tropical peatlands and better informs predictions as the climate changes.

Recent advances in Artificial Intelligence (AI) have brought AI closer to laypeople than ever before. This leads to a pervasive problem: how would a user ascertain whether an AI system will be safe, reliable, or useful in a given situation? This problem becomes particularly challenging when it is considered that most autonomous systems are not designed by their users; the internal software of these systems may be unavailable or difficult to understand; and the functionality of these systems may even change from initial specifications as a result of learning. To overcome these challenges, this dissertation proposes a paradigm for third-party autonomous assessment of black-box taskable AI systems. The four main desiderata of such assessment systems are: (i) interpretability: generating a description of the AI system's functionality in a language that the target user can understand; (ii) correctness: ensuring that the description of AI system's working is accurate; (iii) generalizability creating a solution approach that works well for different types of AI systems; and (iv) minimal requirements: creating an assessment system that does not place complex requirements on AI systems to support the third-party assessment, otherwise the manufacturers of AI system's might not support such an assessment. To satisfy these properties, this dissertation presents algorithms and requirements that would enable user-aligned autonomous assessment that helps the user understand the limits of a black-box AI system's safe operability. This dissertation proposes a personalized AI assessment module that discovers the high-level ``capabilities'' of an AI system with arbitrary internal planning algorithms/policies and learns an accurate symbolic description of these capabilities in terms of concepts that a user understands. Furthermore, the dissertation includes the associated theoretical results and the empirical evaluations. The results show that (i) a primitive query-response interface can enable the development of autonomous assessment modules that can derive a causally accurate user-interpretable model of the system's capabilities efficiently, and (ii) such descriptions are easier to understand and reason with for the users than the agent's primitive actions.

Covalent organic frameworks (COFs) are a recently discovered class of nanoporous polymeric materials with ultra-high specific surface areas, which makes them highly attractive for applications in nanofiltration, gas capture and storage, and catalysis. However, the macroscopic morphology of COFs is maintained by relatively weak physical interactions between crystallites, which limits the applications of COFs where they may experience significant physical stresses. Herein, fillers are added to three-dimensional TAPB-PDA COF aerogels synthesized to improve the mechanical strength and functionality through the formation of a composite material by physically implanting the fillers in the macropores present in the gel network. Boron nitride loading is shown to double the Young’s modulus of the aerogel, from 11 kPa to 22 kPa, at 20 relative weight percent loading, while only causing a 10% decrease in accessible nanoporous surface area, normalized to the mass of COF in the sample. Poly(acrylic acid) added at 5 relative weight percent loading and crosslinked increases the Young’s modulus to 21 kPa and simultaneously increases the elastic limit of the aerogel from 10% to 65% strain, while inducing a 38% decrease in accessible nanoporous surface area, normalized to the mass of COF in the sample. This work demonstrates the potential for macroscopic composites with COFs forming the majority phase of the material, showing the possibility for mechanical reinforcement without significant hinderance of the adsorbent functionality of the material.

This dissertation contributes to uncertainty-aware neural networks using multi-modality data, with a focus on industrial and aviation applications. Drawing from seminal works in recent years that have significantly advanced the field, this dissertation develops techniques for incorporating uncertainty estimation and leveraging multi-modality information into neural networks for tasks such as fault detection and environmental perception. The escalating complexity of data in engineering contexts demands models that predict accurately and quantify uncertainty in these predictions. The methods proposed in this document utilize various techniques, including Bayesian Deep Learning, multi-task regularization and feature fusion, and efficient use of unlabeled data. Popular methods of uncertainty quantification are analyzed empirically to derive important insights on their use in real world engineering problems. The primary objective is to develop and refine Bayesian neural network models for enhanced predictive accuracy and decision support in engineering. This involves exploring novel architectures, regularization methods, and data fusion techniques. Significant attention is given to data handling challenges in deep learning, particularly in the context of quality inspection systems. The research integrates deep learning with vision systems for engineering risk assessment and decision support tasks, and introduces two novel benchmark datasets designed for semantic segmentation and classification tasks. Additionally, the dissertation delves into RGB-Depth data fusion for pipeline defect detection and the use of semi-supervised learning algorithms for manufacturing inspection tasks with imaging data. The dissertation contributes to bridging the gap between advanced statistical methods and practical engineering applications.

My approach to performing contemporary music, like many others, is centeredaround storytelling that merges the intention of the composer with my own interpretation. The balance between the two is unique for every work, as well as the sources of inspiration that shape each interpretation. In some cases, it works well to rely heavily on the historical context of the piece and the specific inspiration and process of the composer. In other cases, the composer desires more freedom and flexibility in the performance of their work, and the story of the piece is woven from the threads of personal stories, emotions, and imagery of the performer. For this project, I made audio recordings of three pieces including Seare Farhat’s Three Children’s Songs for Singing Percussionist, Keiko Abe’s Marimba d’Amore, and Thomas Kotcheffs Obbligato Snare Drum Music No. 1: The Power of Love. I then used these recordings to make music videos that blend elements of pop music videos and classical performance videos, using performance footage as well as narrative and abstract visuals to experiment with video making as a creative outlet while building a performance portfolio that represents me as an artist. In addition to a reflection of my process, this document is also designed as a resource for performers who are interested in learning how to make their own audio and video recordings, covering topics including project planning and preparation, working with collaborators, selecting gear, practicing for studio recordings, and designing and producing videos.

Animal pose estimation (APE) is utilized in preclinical research settings for various neurological disorders such as Parkinson's disease (PD), Huntington's disease (HD) and multiple sclerosis. The technique includes real-time scoring of impairment in the animals during testing or video recording. This is a time-consuming operation prone to errors due to visual fatigue. To overcome these shortcomings, APE automation by deep learning has been studied. The field of APE has gone through significant development backed by improvements in deep learning techniques. These developments have improved 2D and 3D pose estimation, 3D mesh reconstruction and behavior prediction capabilities. As a result, there are numerous sophisticated tools and datasets available today. Despite these developments, APE still lags behind human observer scoring with respect to accuracy and flexibility under complex scenarios. In this project, two critical challenges are being addressed within the context of neurological research focusing on PD. The first challenge is about the lack of comprehensive diverse datasets necessary for accurate training as well as for fine-tuning deep learning models. This is compounded by the inherent difficulty in working with uncooperative rodent subjects, whose unpredictable behaviors often impede reliable data collection. The second challenge focuses on reduction in variation of scores that result from being scored by different evaluators. This will also involve tackling bias and reducing human error for the purpose of reliable and accurate assessments. In order to address these issues, systematic data collection and deep learning in APE have been utilized to automate manual scoring procedures. This project will contribute to neurological research, particularly in understanding and treating disorders like PD. The goal is to improve methods used in assessing rodent behavior which could aid in developing effective therapeutics. The successful implementation of an automated scoring mechanism could set a new standard in neurological research, offering insights and methodologies that are more accurate and reliable.

Phenolic polymers like polyphenols and polyphenylenes have several industrial applications including electrical insulation, specialty membranes, and packings but are typically synthesized under harsh reaction conditions and require hazardous chemicals like formaldehyde. Hydroxycinnamic acids, such as p-coumaric acid (p-CA), are aromatic derivatives of lignin hydrolysates, an underutilized and promising renewable feedstock for production of phenolics and phenolic polymers. Recently a strain of Corynebacterium glutamicum has been created by the Joint BioEnergy Institute (JBEI) which expresses phenolic acid decarboxylase (PAD), an enzyme which catalyzes the reaction of p-CA to 4-vinylphenol (4-VP). Further, a deletion of the phdA gene prevents assimilation of p-CA, thereby increasing 4-VP yield. 4-VP is a substituted phenol which can be polymerized to poly(4-vinylphenol) (PVP) in the presence of ligninolytic enzymes like laccases or peroxidases. This work explores in situ polymerization of 4-VP to PVP by supplementing ligninolytic enzymes during fermentation. Cultured in the presence of p-CA, the engineered C. glutamicum strain achieved a maximum 4-VP yield of 45.2%, 57.9%, and 34.7% when fed 2, 5, and 10 g/L p-CA, respectively. Low yield can be attributed to photodegradation of 4-VP and accumulation of the native laccase present in C. glutamicum which may form only dimers and trimers. To further investigate carbon utilization in the cell, the engineered strain was plasmid cured thus removing the PAD enzyme and fermentations for 13C pathway analysis was performed. Polymerization experiments were performed and the polymer was characterized using GPC.

Coral reefs provide essential social, economic, and ecological services for millions of people worldwide. Yet, climate change and local anthropogenic stressors are damaging reefs globally, compromising reef-building capacity, and therefore impacting functionality. Growth of coral reefs depends upon the production and maintenance of the reef framework when calcium carbonate production exceeds erosion, and utilization of remote sensing to scale-up estimates of reef carbonate production remains limited. This study provided a first field estimate of net carbonate production on Hawaiʻi Island, in Hōnaunau Bay, and used high-resolution benthic-cover data, derived from Global Airborne Observatory (GAO) airborne imaging spectroscopy, to scale-up estimates. Net carbonate production was, on average, 0.5 kg CaCO3 m-2 y-1 across the depth gradient, with the highest rates of approximately 2.4 kg CaCO3 m-2 y-1 at 6 m. Urchins, especially the abundant Echinometra, suppressed reef-accretion potential in the shallow reef (< 6 m) and urchin bioerosion decreased with depth. Critically, a threshold of ~26% live-coral cover is currently needed to maintain positive net production across depths. Scaling-up estimates were achieved using a 2 m resolution map of live-coral cover collected by the GAO. Overall, field measurements translate to average vertical reef growth of 0.5 mm y-1 across depths, whereas sea level is currently increasing at 3.55 mm y-1, suggesting the reef in its present status is not keeping pace with sea-level rise. This work lays the foundation to enhance monitoring of carbonate production over increased temporal and spatial scales with airborne imaging spectroscopy — to help determine where reefs are potentially keeping up with anthropogenic stressors, ocean warming, and sea-level rise — and to help inform restoration and management decisions that support resilient carbonate budgets of coral reefs.