OIST Institutional Repository
Not a member yet
2744 research outputs found
Sort by
Catalytic Enantioselective Mannich and retro-Mannich Reactions and Combinations of These Reactions to Synthesize α,α-Disubstituted α-Amino Acid Derivatives
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of Philosophyα,α-Disubstituted α-amino acid derivatives are found in biofunctional molecules and their building blocks. Whereas various synthetic strategies to access to amino acids derivatives have been reported, synthesis of highly enantiomerically enriched α,α-disubstituted α-amino acid derivatives is still a challenge. In this thesis study, catalytic enantioselective Mannich reactions of ketones with cyclic ketimino esters and kinetic resolutions of the racemic versions of Mannich reaction products via retro-Mannich reactions have been developed to afford α,α-disubstituted α-amino acid derivatives. Further, a strategy that combines the Mannich reaction and the retro-Mannich reaction and a strategy that combines the kinetic resolution with the retro-Mannich reaction have been developed to afford α,α-disubstitutedα-amino acid derivatives with very high enantiopurities (er > 99:1)
Response Properties of Neurons in the Higher Auditory Cortex of Female Zebra Finches, Taeniopygia guttata
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyFemale zebra finches recognize individuals by songs and show a robust preference for their father’s songs (FS) over unfamiliar songs. Using AAV-mediated neuronal ablation, we found that the higher auditory regions, caudomedial nidopallium (NCM) and caudomedial mesopallium (CMM), were necessary for song preference to father’s song (FS) in female adults. However, whether the NCM and/or CMM neurons of females contain the traits of auditory memory for FS has yet to be clarified. I found that NCM neurons responded to a variety of zebra finch songs with only a small fraction of FS-selective neurons. I further found that NCM neurons exhibited selective responses to song element types, but not specifically to the types included in the FS. On the other hand, CMM neurons significantly increased or decreased their firings in response to songs, but were not biased to FS, either. Taken together, these results suggest that both NCM and CMM neurons of female zebra finches did not hold memory traits of FS in the form of single selective neurons. Therefore, the decrease of FS preference with our neuronal ablation might be attributed to the compromise of song discriminability by losing fractions of NCM or CMM neurons
Environmental DNA Metabarcoding for Monitoring Marine Fish Assemblages and Assessing “Tropicalizing” Range Shifts in Japan
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyMarine biodiversity is redistributing due to climate change, necessitating the development of effective biomonitoring tools to understand and potentially mitigate the impacts. Environmental DNA (eDNA) metabarcoding is a highly scalable and non-invasive biodiversity survey method with the potential to accumulate large amounts of both assemblage and species occurrence data. In this thesis, I applied eDNA metabarcoding to analyze the marine fish assemblages and species occurrences across Japan and discuss the findings in comparison to the syntheses of ichthyological literature and reference databases, assessing the strengths and limitations of this method.
A local-scale eDNA survey of the Ogasawara Islands across short intra-island group distances detected higher prevalence and relative abundance of the Scleractinia coral genera Acropora spp. in disagreement with past reports and provided novel occurrences for fish species and coral genera, representing a successful use-case of eDNA for efficient multi-taxa surveys of isolated, data-deficient ecosystems (Chapter 2). Consolidating multiple collaborative surveys and open access sequencing datasets resulted in possibly the largest fish eDNA dataset to date for Japan, enabling a nation-wide comparative assessment of sample completeness and fish assemblage diversity and evenness standardized by sampling coverage (Chapter 3). Most of the highly frequent species were detected across varying biogeographical zones and the obtained diversity patterns were mostly consistent with literature albeit with caveats. Mostly reef-associated or benthic Actinopterygii were detected, reflecting the sampling scheme almost exclusive to nearshore surfaces. Finally (Chapter 4), analysis of the eDNA occurrence records in comparison to the historical occurrence records and northern limit estimates from Aquamaps species distribution models and IUCN expert range maps showed that eDNA can detect high numbers of spatiotemporally resolved occurrences from marginal range edges: of the total 36,282 eDNA occurrence records, 1,337 (3.7%) were marginal for 193 species with regards to IUCN and 2,410 (6.6%) were marginal for 267 species with regards to Aquamaps. eDNA-sampling obtained novel latitudinal records and helped identify potentially outdated reference distribution ranges and limits for various commercially important species (e.g. Decapterus maruadsi, Epinephelus fasciatus, Pterocaesio spp.), cryptic taxa (e.g. Entomacrodus striatus, Enneapterygius philippinus), herbivores implicated in “isoyake” barrens (e.g. Siganus fuscescens, Calotomus japonicus, Prionurus scalprum) and overwintering seasonal vagrants (e.g. Ctenochaetus striatus, Sargocentron praslin, Acanthurus lineatus).
Overall, this series of studies demonstrates the power and limitations of eDNA metabarcoding as a long-term tool for marine fish diversity assessment and tracking range extensions. Its scalability enables high spatiotemporal coverage to reveal local and national-scale biodiversity patterns (Chapter 2 and 3) and possibly better track species distribution range extensions contributing to tropicalization (Chapter 4). However, integration with traditional surveys and contextualization of the findings with literature and expert knowledge are key to interpretations and optimization of this method that is here to stay
A Neural Network Model Inspired by the Prefrontal-Thalamic-Hippocampal Circuits for Navigation under Complex Spatio-temporal Rules
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyI investigated the functional relationship between the brain regions related to contextdependent memory and aimed to find features of their interactions, particularly information flows among these regions, through brain function modeling using machine learning techniques. I focused on the interaction between the Hippocampus(HPC), Prefrontal cortex(PFC), and nucleus reuniens (Re), which is a component of the thalamus, for the retrieval of contextual information. The modules describing these regions are built using bidirectionally interconnected long-short-term memory (LSTM) units. My results highlight the critical roles of distinct brain modules. These roles and functional interplay enhance the robustness of the system in handling complex cognitive tasks. Furthermore, I simulated a similar network model with reinforcement learning and demonstrated the advantages of the multi-module network structure in rapid adaptation to changes in contextual information over the single-module network structure. In sum, the proposed model inspired by the PFC-ReHPC neural circuit demonstrates robustness and flexibility in learning navigation in complex environments, suggesting the functional benefits of the multi-module network
Investigations into the Neuroscience of the Self: Personality and Age from the Perspective of Intrinsic Neural Timescales
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyThe Pattern Theory of Self (PTS) proposes that selfhood emerges from dynamic patterns of interrelated factors anchored in a living body, encompassing elements such as bodily processes, affective states, social interactions, and cognitive processes. Northoff’s temporospatial theory suggests that the brain’s intrinsic temporospatial dynamics constitute the neural foundation of self-experiences. Empirical evidence has demonstrated that longer neural timescales, measured via EEG autocorrelation windows (ACWs), are associated with self-related processing.
However, the relationship between these temporal dynamics and individual characteristics remains unclear. This thesis presents four studies across two datasets to investigate the relationships between the behavioral, neural, and psychological components of the self-pattern.
Using a Perceptual Crossing Experiment (PCE) paradigm involving real-time interpersonal interaction, Studies 1-2 examined personality-behavior relationships and age-ACW associations in a predominantly young adult sample. Using the Leipzig Mind-Brain-Body (LEMON) dataset, Studies 3-4 investigated age-ACW relationships across the adult lifespan and examined the connections between personality traits, neural timescales, and spontaneous thought patterns. The results revealed robust age effects on neural timescales, with older adults showing shorter ACWs (d = 0.38-0.51). However, personality traits showed minimal relationships with behavior during social interactions, no associations with neural timescales, and only selective relationships with thought patterns (primarily Neuroticism with negative/temporal thinking). These findings highlight the complex relationships between the different components of self-patterns
Investigating the Neural and Behavioural Features of Flexible Task Performance in the Olfactory Working Memory Task
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyBehaviour arises from the transformation of sensory inputs into context-appropriate actions, a process shaped by relationships among stimuli. Perceptual similarity, in particular, may constrain how easily new associations are learned or flexibly updated. This thesis examines how perceptual similarity shapes odour-guided behaviour and its cortical representations during an olfactory delayed non-match to sample (DNMS) task.
Using modified versions of the DNMS task, I first investigate how stimulus timing and odour similarity influence the initial acquisition of this paradigm. High odour similarity slowed learning, consistent with the idea that overlapping sensory representations make early discriminations more difficult. A comparable effect emerged during rule switches, where closely related odours slowed adjustment to changing contingencies.
To relate these behavioural effects to cortical activity, I recorded population responses across higher cortical regions using Neuropixel probes. The orbitofrontal cortex preserved odour identity information and captured relationships between similar stimuli. Premotor regions showed only limited odour information but carried signals related to whether the odours matched or differed during the decision period. These observations indicate that different higher cortical areas emphasise distinct aspects of the task, from representing perceptual features to signalling trial outcomes.
Together, the behavioural and physiological findings show how perceptual similarity shapes the acquisition and flexible use of learned associations, and how distinct cortical areas emphasise different aspects of task-relevant informatio
Learning to Be Creative: A Computational Investigation of the Role of Associative Learning in Creativity
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyThe Self-Optimization (SO) model can be considered the third operational mode of the classical Hopfield Network (HN). Through associative learning within its own state space it leverages the power of associative memory to optimize its own behavior without any external rewards. It is an elegant mathematical model, with a simple unsupervised Hebbian learning rule, applicable to a wide range of complex adaptive systems. In this thesis, we propose the SO model as a minimal model of creativity and a framework, which allows us to examine the relationship between creativity and associative learning.
We examine the existing computational investigations into minimal creativity via a bibliometric analysis of the creativity literature of the past century. Our analysis highlights a theoretical research gap - there is little work on minimal creativity, let alone discussion of computational approaches to modeling it.
We then evaluate the SO model as a candidate for investigating minimal creativity based on its computational elegance, biological plausibility, and relationship to minimal forms of agency in living organisms. We analyze how the workings of the SO model align with the product- and process-based perspectives of creativity. We determine that the SO model meets all the requirements to constitute a creative process and to yield creative products as solutions to its optimization process. Crucially, we show that creative products as a result of learning are generated with above chance probability.
We extend the biological plausibility aspect of the SO model by applying partial system resets. We also show that this may be more temporally and computationally efficient than doing full resets of the entire system. We enhance the engineering aspect of the model by developing a new implementation that substantially reduces its computational effort, enabling the investigation of dramatically larger systems.
Finally, by translating combinatorial propositional satisfiability problems (SAT) into the HN weights, we demonstrate that the SO model can be applied to SAT problems. This not only makes the SO model relevant to various scientific fields, where real-world problems can be expressed as SAT problems, but also allows us to examine, perhaps for the first time, the relationship between associative learning and constraints of the problem. We discuss why investigating concrete problems is imperative for a more informed investigation of learning effects on the spectrum of possible creative outcomes
An Investigation on Lie Algebraic Aspect of Quantum Fisher Information Applied to Unitary and Non-Unitary Evolution
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyIn this thesis, I focus on the quantum Fisher information (QFI) - a central quantity in parameter estimation theory for quantum systems, which provides a lower bound on the precision one can achieve. In the first part of the thesis, I explore a Lie algebraic approach to evaluate the quantum Fisher information for unitary dynamics. I show how the Lie algebra associated with the unitary evolution, together with the initial state, allows one to compute various quantities of interest, circumventing the computation of the time-evolved state. In particular, I apply this shortcut approach to calculate the QFI and use it for the QFI evaluation in various prototype models, both finiteand infinite-dimensional. In the second part of the thesis, I shift focus to non-unitary dynamics, such as those described by a quantum master equation. In this setting, I theoretically study the description of open system dynamics in the Liouville space of vectorized density matrices and investigate the so-called dissipative quantum Fisher information (DQFI) within that framework. Specifically, I demonstrate how the metric defined in Liouville space differs from its counterpart in Hilbert space. I establish a map between the Hilbert space QFI and the Liouville dissipative quantum Fisher information for qubit and qudit systems, expressed via the quantum state’s purity. Furthermore, I discuss the physical meaning of the DQFI by studying its behavior under completely positive trace-preserving (CPTP) maps. The present results demonstrate that the Lie algebraic approach provides a computationally efficient framework for the QFI evaluation. This approach allows to avoid the explicit solution of quantum dynamics, and remains well-defined in cases when conventional approaches like WeiNorman expansion exhibit singularities
Behavioural and Neurobiological Mechanisms Underlying Different Navigational Strategies
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyNavigation is fundamental for all animals, whether to find food, shelter, or a mate. The study of navigation is necessary not only because it is crucial for survival, but also because it is an intricate behaviour that involves complex cognitive functions such as planning, decision-making, and memory. When navigating, animals can rely on external environmental cues (allocentric navigation) or internal cues such as body position and movement (egocentric navigation). Understanding why animals use each strategy and investigating how place cells represent spatial information under them can deepen our understanding of how animals make sense of the physical world. Studying these different navigational systems can also provide a window into understanding the interaction between different memory systems. This work aims to (1) identify environmental and behavioural factors that lead animals to use different navigational strategies; and (2) compare place cell properties and stability in mice using an allocentric or egocentric navigational strategy. In the first part of this work, I manipulated different training and environmental variables to elucidate which factors promote allocentric versus egocentric strategy usage. I found that in a plus-shaped maze, mice predominantly used the egocentric strategy, regardless of cue availability or the duration of training. The only factor that led to the usage of an allocentric strategy was the shape of the maze itself. In the second part of this work, I compared place cell properties between the two strategies by recording calcium activity of pyramidal cells in the CA1 of mice navigating with an allocentric or egocentric strategy. I found that both allocentric and egocentric navigation are encoded in the hippocampus of these mice and that both strategies are similarly stably represented
The Evolution and Functional Morphology of the Ant Mesosoma
Full text linkOkinawa Institute of Science and Technology Graduate UniversityDoctor of PhilosophyComparative functional morphology seeks to understand the patterns and mechanisms that shape the diversity of life by examining variation among species. We have much to learn about the engineering principles of nature and how these shape evolution over time. Ants, with their taxonomic and morphological diversity as well as ecological dominance, are an excellent model for exploring the link between form and function. Given the role of mesosoma in supporting body weight in ants, enabling locomotion, and other functions such as carrying prey, we hypothesized that as ants diversified into different niches, the skeletomuscular system of mesosoma was modified to fit for different ecological roles. To test this hypothesis, I pursued three subprojects: In the first chapter, I performed a detailed morphological analysis of the skeletomuscular system of the mesosoma of ecologically generalized ant Formica rufa. These analyses provide a comprehensive atlas of the skeletomuscular system of the ant mesosoma and provide a functional interpretation of the muscles. In the second chapter, I focused on a specialized mode of locomotion in ants: forward leaping. I investigated the morphological changes associated with jumping behavior by comparing the musculature between jumping and non-jumping ants. I demonstrated that the enlargement of the depressor muscles is the primary factor underlying the jumping ability in ants. Finally, in the third chapter, I studied the evolutionary patterns of mesosoma in ants using phylogenetic comparative methods. Our data suggest that size-related changes, correlated evolution of traits, and phylogeny play a significant role in explaining the variability in mesosoma muscles, however, there was no effect of the ecological niche. This dissertation is the first step towards understanding the evolutionary transformations and functional and ecological diversification of mesosoma across Formicidae