Biology subjects

Chung, U.-i.

Publications and source records attributed to Chung, U.-i..

8 recordsLinked to original sources

bioRxiv2026

Collective AI use is associated with researcher engagement: Real-time evidence from a scientific conference

Recent large-scale bibliometric analyses suggest that individual AI use can increase productivity while reducing downstream engagement and topic diversity. Here we ask whether collective AI deployment is associated with shared engagement. Using an Audience Response Engagement (ARE) system at NGS Expo 2025 (N=110 biomedical researchers), we captured real-time consensus and generated updated visualizations within minutes. Our data reveal a substantial gap between adoption and transparency: 93.6% of researchers use AI at least weekly, yet only 5.5% consistently disclose this usage--a 17-fold disparity. This pattern is consistent with systemic policy uncertainty (39.1% report unclear guidelines). Behavioral clustering identifies a "High-Concern" group (31%) as a candidate for targeted interventions: highest productivity yet lowest disclosure. These findings suggest that collective AI deployment in physical settings is associated with shared engagement.

scientific communication and education↗

bioRxiv2024

Modeling of vascular interactions in endochondral ossification using human embryonic stem cells-derived organoid on a microfluidic chip

Vascular interactions play a crucial role in embryogenesis, including skeletal development. During endochondral ossification, vascular networks are formed as mesenchymal cells condense and later invade skeletal elements to form the bone marrow. We and other groups developed a model of endochondral ossification by implanting human embryonic stem cell (hESC)-derived sclerotome into immunodeficient mice. However, in vitro models of endochondral ossification, particularly vascular interaction with mesenchymal cells at its initial stage, are yet to be established. Therefore, we developed a method to model the initial stage of endochondral ossification using a microfluidic chip-based platform, with a particular focus on the vascular interaction. On the chip, we found that the fibrin gel helped align mCherry-expressing human umbilical vein endothelial cells (HUVECs) better than the collagen-I gel, suggesting that the fibrin gel is more suitable for the formation of a vascular-like network. The perfusability of the vascular-like networks was partially confirmed using fluorescein isothiocyanate (FITC)-dextran and fluorescent microbeads. We then mixed hESC-derived sclerotome with enhanced green fluorescent protein (EGFP)-expressing HUVECs and applied this mixture on the chip. We named this mixture of cells SH organoids. The SH organoids showed superior abilities to maintain the vascular-like network, which was formed by the mCherry-expressing HUVECs, compared with the sclerotome spheroids on the chip. The EGFP-expressing HUVECs migrated from the SH organoid, formed a vascular-like networks, and partially interacted with the mCherry-expressing vascular-like networks on the chip. Histological analysis showed that SRY-box transcription factor 9 (SOX9) and type I collagen were expressed mutually exclusively in the condensed mesenchymal cells and perichondrial-like cells, respectively. This study demonstrates that our SH organoid-on-a-chip method reproduces vascular networks that are formed at the initial stage of endochondral ossification. This model may provide insights into human endochondral ossification and has potential applications in bone disease modeling and drug screening.

bioengineering↗

bioRxiv2022

intra-single cell sequencing (iSCseq) spotlights transcriptomic and epigenetic heterogeneity inside multinucleated osteoclast

Single-cell RNA-seq (scRNA-seq) has clarified cellular heterogeneity within cell populations. However, scRNA-seq and spatial transcriptomics cannot capture the dynamic transcriptomic changes inside living cells. To decode subcellular gene expression, we developed intra-single cell sequencing (iSCseq), a novel approach that combines confocal imaging, repeatedly picking up cellular components inside living cells, and next-generation sequencing (intra single-cell RNA-seq; iSCseq). iSCseq illustrated the subcellular heterogeneity of gene expression. iSCseq revealed not only multiple differentiation stages embedded in the same cell, but also physical cytoskeletal connections, physiological activity of mitochondria, and intracellular calcium, as confirmed by transcriptomic evidence. Inclusive iSCseq with in vivo scRNA-seq datasets identified new osteoclast subsets in physiological and pathological bones. Network analysis with centrality provided insights into the connection between subcellular components, and clearly divided differentiation and fusion processes in multinucleation. The iSCseq approach has the potential to enhance cell biology at subcellular resolution and identify new therapeutic targets. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=200 SRC="FIGDIR/small/506360v4_ufig1.gif" ALT="Figure 1"> View larger version (73K): org.highwire.dtl.DTLVardef@115c1d8org.highwire.dtl.DTLVardef@1695ad2org.highwire.dtl.DTLVardef@1c9d257org.highwire.dtl.DTLVardef@1107ddb_HPS_FORMAT_FIGEXP M_FIG C_FIG In briefintra-single cell sequencing (iSCseq) enhances single-cell technology by combining live cell imaging, subcellular sampling from living cells and sequencing, offering deeper insights into cell functions and pathology at subcellular resolution through inclusive analysis with scRNA-seq and advanced centrality-focused network analysis. HighlightsO_LIintra-single cell sequencing (iSCseq) clarifies subcellular heterogeneity C_LIO_LIiSCseq connects morphological and physiological features with transcriptome C_LIO_LIInclusive iSCseq unveils osteoclast subsets in physiological and pathological bones C_LIO_LILinkage at subcellular resolution reveals key players in characteristic fusion C_LI

genomics↗

bioRxiv2022

Injectable hydrogels with phase-separated structures that can encapsulate live cells

Injectable hydrogels are biomaterials that can be administered minimally invasively in liquid form, and are considered promising artificial extracellular matrix (ECM) materials. However, ordinary injectable hydrogels are synthesized from water-soluble molecules to ensure injectability, resulting in nonphase-separated structures, making them structurally different from natural ECMs with phase-separated insoluble structural proteins, such as collagen and elastin. Here, we propose a simple material design approach to impart phase-separated structures to injectable hydrogels by adding inorganic salts. Injecting a gelling solution of mutually crosslinkable tetra-arm poly(ethylene glycol)s with potassium sulfate at optimal concentrations results in the formation of a hydrogel with phase-separated structures in situ. These phase-separated structures provide up to an 8-fold increase in fracture toughness while allowing the encapsulation of live mouse chondrogenic cells without compromising their proliferative activity. Our findings highlight that the concentration of inorganic salts is an important design parameter in injectable hydrogels for artificial ECMs.

bioengineering↗

bioRxiv2021

A generalized distribution interpolated between the exponential and power law distributions and applied to the walking data of the pill bug (Armadillidium vulgare)

The Levy walk, a type of random walk in which the frequency of linear step-lengths follows a power-law distribution, can be observed in the migratory behavior of organisms at various levels, from bacteria and T cells to humans. Compared to the Brownian walk, which is also a type of random walk (characterized by an exponential distribution of the frequency of occurrence of step-length), the Levy walk is characterized by the occasional appearance of linear movements over very long distances. In this paper, we propose a general distribution that includes the power-law and exponential distributions as special cases. This distribution has two parameters: the first parameter represents the exponent, similar to the power-law and exponential distributions and the second is a shape parameter representing the shape of the distribution. By introducing this distribution, an intermediate distribution model can be interpolated between the power-law and exponential distributions. The shape parameter measures whether a distribution assimilates a power-law or exponential distribution. In this study, the proposed distribution was fitted to the frequency distribution of the step-length calculated from the walking data of pill bugs. The autocorrelation coefficients were also calculated from the time-series data of the step-length, and the relationship between the shape parameter and time dependency was investigated. The results showed a significant negative correlation between the two (r=-0.61, n=30, t=4.04, and p=0.00037). This means that pill bugs with gait patterns closer to Levy than Brownian walks have a stronger time dependence with respect to step-length changes.

animal behavior and cognition↗

bioRxiv2021

Simulation of foraging behavior using a decision-making agent with Bayesian and inverse Bayesian inference: Levy and Brownian walk-like search patterns derived from differences in prey distributions

It has been stated that in human migratory behavior, the step length series may have temporal correlation and that there is some relationship between this time dependency and the fact that the frequency distribution of step length follows a power-law distribution. Furthermore, the frequency of occurrence of the step length in some large marine organisms has been found to switch between power-law and exponential distributions, depending on the difficulty of prey acquisition. However, to date it has not been clarified how the aforementioned three phenomena arise: the positive correlation created in the step length series, the relation between the positive correlation of the step length series and the form of an individuals step length distribution, and the switching between power-law and exponential distributions depending on the abundance of prey. This study simulated foraging behavior using the Bayesian decision-making agent simultaneously performing both knowledge learning and knowledge-based inference to analyze how the aforementioned three phenomena arise. In the agent with learning and inference, past experiences were stored as hypotheses (knowledge) and they were used in current foraging behavior; at the same time, the hypothesis continued to be updated based on new experiences. The simulation results show that the agent with both learning and inference has a mechanism that simultaneously causes all the phenomena.

animal behavior and cognition↗

bioRxiv2019

A role of causal inference in non-stationary environments

Bayesian inference is the process of narrowing down the hypotheses (causes) to the one that best explains the observational data (effects). To accurately estimate a cause, a considerable amount of data is required to be observed for as long as possible. However, the object of inference is not always constant. In this case, a method such as exponential moving average (EMA) with a discounting rate is used to improve the ability to respond to a sudden change; it is also necessary to increase the discounting rate. That is, a trade-off is established in which the followability is improved by increasing the discounting rate, but the accuracy is reduced. Here, we propose an extended Bayesian inference (EBI), wherein human-like causal inference is incorporated. We show that both the learning and forgetting effects are introduced into Bayesian inference by incorporating the causal inference. We evaluate the estimation performance of the EBI through the learning task of a dynamically changing Gaussian mixture model. In the evaluation, the EBI performance is compared with those of the EMA and a sequential discounting expectation-maximization algorithm. The EBI was shown to modify the trade-off observed in the EMA.

animal behavior and cognition↗

bioRxiv2019

Extended Bayesian inference incorporating symmetry bias

In this study, we start by proposing a causal induction model that incorporates symmetry bias. This model has two parameters that control the strength of symmetry bias and includes conditional probability and conventional models of causal induction as special cases. It can reproduce causal induction of human judgment with high accuracy. We further propose a human-like Bayesian inference method to replace the conditional probability in Bayesian inference with the aforementioned causal induction model. In this method, two components coexist: the component of Bayesian inference, which updates the degree of confidence for each hypothesis, and the component of inverse Bayesian inference that modifies the model of each hypothesis. In other words, this method allows not only inference but also simultaneous learning. Our study demonstrates that the method with both Bayesian inference and inverse Bayesian inference enables us to deal flexibly with unsteady situations where the target of inference changes occasionally.

animal behavior and cognition↗