Biology subjects

Aihara, K.

Publications and source records attributed to Aihara, K..

14 recordsLinked to original sources

bioRxiv2026

Dynamical Biomarkers of Creative Cognition Across Divergent and Convergent Problem-Solving

Creativity fluctuates markedly from moment to moment, even within the same individual, yet the neural dynamics that determines whether a given attempt produces a highly creative idea remains poorly understood. Prior studies have identified static EEG correlates of creative thinking, but these do not explain how brain activity is dynamically organized before and during successful idea generation. Here, we model creative cognition as trajectories through a neural state space using energy landscape analysis (ELA) of EEG recorded during two complementary problem-solving paradigms: the divergent Alternative Uses Test (AUT) and the convergent, goal-directed Fusion Innovation Test (FIT). Across both tasks, creative success was associated with dissociable dynamical signatures in the resting and ideation stages. Before ideation, greater diversity of resting-state patterns of activity, corresponding to possible attractors, indexed a preparatory substrate of creative potential, showing weak trial-level effects but robust subject-level coupling with performance. During ideation, higher creativity was predicted by how the brain traversed its accessible state space: successful trials were characterized by traversal biased toward sustained exploration within stable attractor basins rather than frequent switching between basins of attraction ({beta} = 0.104, p = 0.004, trial-level). These findings identify a task-invariant, biologically grounded dynamical mechanism of creative cognition and show that creative performance depends not only on which neural states are available, but also on how neural activity traverses that state space.

neuroscience↗

bioRxiv2026

Controlling Integration and Segregation in Echo State Networks via Noradrenaline and Acetylcholine Neuromodulation

Biological brain networks flexibly reconfigure functional connectivity between integration and segregation through neuromodulatory systems--noradrenaline (NA) and acetylcholine (ACh)--without altering structural connectivity. Inspired by this mechanism, we propose a modular echo state network (ESN) with context-dependent NA and ACh gain modulation, where NA promotes inter-module integration via response gain and ACh promotes intra-module segregation via multiplicative gain. We evaluate the model on two context-dependent tasks: a segregation/integration task and a context-dependent decision task. Across both tasks, the modulated model consistently outperformed the baseline, with task-appropriate modulation profiles emerging naturally from optimization and functional connectivity analysis confirming context-appropriate dynamic reorganization. These results demonstrate that neuromodulatory gain control enables adaptive, context-sensitive computation in structurally fixed reservoir networks.

neuroscience↗

bioRxiv2025

BrainBeacon: A Cross-Species Foundation Model for Single-cell Resolved Brain Spatial Transcriptomics

The brains functional complexity emerges from spatiotemporal heterogeneity in molecular, cellular, and circuit organization, shaped by evolution, aging, and disease. Single-cell spatial transcriptomics provides a transformative framework to decode this complexity, resolving cell-type-specific expression patterns, region-specific regulation, evolutionary divergence, and disease-associated disruptions. Despite recent advances in spatial transcriptomics, a unified foundation model capturing cross-species, whole-brain cellular architecture remains lacking. Here, we introduce BrainBeacon, the first cross-species foundation model for whole-brain spatial transcriptomics. Trained on 133 million spatially resolved cells covering a total area of 210194 mm2 from human, macaque, marmoset and mouse brains profiled across five major platforms, BrainBeacon integrates gene expression rankings, spatial cellular organization, and evolutionarily conserved genetic relationships into a unified transformer-based architecture. A two-stage training strategy captures both intra-cellular transcriptional dependencies and inter-cellular spatial interactions, yielding biologically grounded and interpretable representations of brain structure. Subsequently, BrainBeacon enables cross-species alignment of brain cell types and anatomical regions, facilitating integrative analyses of conserved and divergent cellular architectures. Moreover, it uncovers spatial regulatory mechanisms of aging by predicting bidirectional perturbation responses between cells and their local niche. Together, BrainBeacon serves as a foundational cross-species model for whole-brain spatial transcriptomics, enabling virtual brain atlas construction, systematic in silico perturbation studies, and mechanistic insights into brain disease pathogenesis.

bioinformatics↗

bioRxiv2025

Longitudinal analysis of antibody titers after primary and booster mRNA COVID-19 vaccination can identify individuals at risk for breakthrough infection

A key issue in the post-COVID-19 era is the ongoing administration of COVID-19 vaccines. Repeated vaccination is essential for preparing against currently circulating and newly emerging SARS-CoV-2 variants while enabling people to continue with daily life. Optimizing vaccination strategies is crucial to efficiently manage medical resources and establish an effective vaccination framework. Therefore, it is important to quantitatively understand vaccine-induced immunity dynamics and to be able to identify poor responders with lower sustained antibody titers as potential priorities for revaccination. We investigated longitudinal antibody titer data in a cohort of 2,526 people in Fukushima, Japan, from April 2021 to November 2022 for whom basic demographic and health information was available. Using mathematical modeling and machine learning, we stratified the time-course patterns of antibody titers after 2 primary doses and 1 booster dose of mRNA COVID-19 vaccines. We identified 3 notable populations, which we refer to as the durable, the vulnerable, and the rapid-decliner populations, approximately half of which remained in the same population after the booster dose. Notably, the rapid-decliner population experienced earlier infections than the others. Furthermore, when comparing IgG(S) titers, IgA(S) titers, and T-spot counts between participants who experienced breakthrough infections after booster vaccination and those who did not, we found that IgA(S) titers were significantly lower in breakthrough infected participants during the early stage after booster vaccination. Our computational approach is adaptable to various types of vaccinations. This flexibility can inform policy decisions on vaccine distribution to enhance immunity both in future pandemics and in the post-COVID-19 era.

immunology↗

bioRxiv2024

Steering cell-state and phenotype transitions by causal disentanglement learning

Understanding and manipulating cell-state and phenotype transitions is essential for advancing biological research and therapeutic treatment. We introduce CauFinder, an advanced framework designed to accurately identify causal regulators of these transitions and further precisely steer such transitions by integrating causal disentanglement modelling with network control based solely on observed data. By leveraging do-calculus and optimizing information flow metrics, CauFinder can distinguish causal factors from spurious ones, ensuring precise control over desired state transitions. One significant advantage of CauFinder is its ability to identify those variables causally affecting the cell-state/phenotype transitions among all observed variables, both theoretically and computationally, leading to the identification of their master regulators when combined with network control. Consequently, by employing a counterfactual algorithm, CauFinder is able effectively to facilitate desirable state transitions or steer these transitional trajectories/paths by modulating these causal drivers. Beyond its theoretical advantages, CauFinder outperforms existing approaches computationally in both simulated and real-world settings. CauFinder is able to not only reveal natural biological transition processes such as (a) cell differentiation, (b) lung adenocarcinoma (LUAD) to lung squamous cell carcinoma (LUSC) transdifferentiation and (c) drug-sensitive to drug-resistant transitions but also identify the causal regulators of their reverse transition processes, such as (A) cell dedifferentiation, (B) LUSC to LUAD transdifferentiation and (C) drug-resistant to drug-sensitive transitions. These findings highlight its superior ability to causally uncover essential regulatory mechanisms and accurately steer cell-state/phenotype transitions, thus providing novel therapeutic strategies.

bioinformatics↗

bioRxiv2024

Stratification of viral shedding patterns in saliva of COVID-19 patients

Living with COVID-19 requires continued vigilance against the spread and emergence of variants of concern (VOCs). Rapid and accurate saliva diagnostic testing, alongside basic public health responses, is a viable option contributing to effective transmission control. Nevertheless, our knowledge regarding the dynamics of SARS-CoV-2 infection in saliva is not as advanced as our understanding of the respiratory tract. Here we analyzed longitudinal viral load data of SARS-CoV-2 in saliva samples from 144 patients with mild COVID-19 (a combination of our collected data and published data). Using a mathematical model, we quantified individual-level viral dynamics and stratified them into three groups using a clustering approach. Notably, the three groups exhibited distinct differences viral RNA detection durations: 11.5 days (95% CI: 10.6 to 12.4), 17.4 days (16.6 to 18.2), and 30.0 days (28.1 to 31.8), respectively. Surprisingly, this stratified grouping remained unexplained despite our analysis of 47 types of clinical data, including basic demographic information, clinical symptoms, results of blood tests, and vital signs. Additionally, we quantified the expression levels of 92 micro-RNAs in a subset of saliva samples, but these also failed to explain the observed stratification, although the mir-1846 level may have been weakly correlated with peak viral load. Our study provides insights into SARS-CoV-2 infection dynamics in saliva, highlighting the challenges in predicting the duration of viral RNA detection without indicators that directly reflect an individuals immune response, such as antibody induction. Given the significant individual heterogeneity in the kinetics of saliva viral shedding, identifying biomarker(s) for viral shedding patterns will be crucial for improving public health interventions in the era of living with COVID-19.

microbiology↗

bioRxiv2023

Multiscale modeling of HBV infection integrating intra- and intercellular viral propagation for analyzing extracellular viral markers

Chronic infection of hepatitis B virus (HBV) is caused by the persistence of closed circular DNA (cccDNA) in the nucleus of infected hepatocytes. Despite available therapeutic anti-HBV agents, eliminating the cccDNA remains challenging. The quantifying and understanding dynamics of cccDNA are essential for developing effective treatment strategies and new drugs. However, it requires a liver biopsy to measure the intrahepatic cccDNA, which is basically not accepted because of the ethical aspect. We here aimed to develop a non-invasive method for quantifying cccDNA in the liver using surrogate markers present in peripheral blood. We constructed a multiscale mathematical model that explicitly incorporates both intracellular and intercellular HBV infection processes. The model, based on age-structured partial differential equations (PDEs), integrates experimental data from in vitro and in vivo investigations. By applying this model, we successfully predicted the amount and dynamics of intrahepatic cccDNA using specific viral markers in serum samples, including HBV DNA, HBsAg, HBeAg, and HBcrAg. Our study represents a significant step towards advancing the understanding of chronic HBV infection. The non-invasive quantification of cccDNA using our proposed methodology holds promise for improving clinical analyses and treatment strategies. By comprehensively describing the interactions of all components involved in HBV infection, our multiscale mathematical model provides a valuable framework for further research and the development of targeted interventions.

microbiology↗

bioRxiv2023

Isolation may select for earlier and higher peak viral load but shorter duration in SARS-CoV-2 evolution

During the COVID-19 pandemic, human behavior change as a result of nonpharmaceutical interventions such as isolation may have induced directional selection for viral evolution. By combining previously published empirical clinical data analysis and multi-level mathematical modeling, we found that the SARS-CoV-2 variants selected for as the virus evolved from the pre-Alpha to the Delta variant had earlier and higher infectious periods but a shorter duration of infection. Selection for increased transmissibility shapes the viral load dynamics, and the isolation measure is likely to be a driver of these evolutionary transitions. In addition, we showed that a decreased incubation period and an increased proportion of asymptomatic infection were also positively selected for as SARS-CoV-2 mutated to the extent that people did not isolate. We demonstrated that the Omicron variants evolved in these ways to adapt to human behavior. The quantitative information and predictions we present here can guide future responses in the potential arms race between pandemic interventions and viral evolution.

evolutionary biology↗

bioRxiv2022

STAMarker: Determining spatial domain-specific variable genes with saliency maps in deep learning

Spatial transcriptomics characterizes gene expression profiles while retaining the information of the spatial context, providing an unprecedented opportunity to understand cellular systems. One of the essential tasks in such data analysis is to determine spatially variable genes (SVGs), which demonstrate spatial expression patterns. Existing methods only consider genes individually and fail to model the inter-dependence of genes. To this end, we present an analytic tool STAMarker for robustly determining spatial domain-specific SVGs with saliency maps in deep learning. STAMarker is a three-stage ensemble framework consisting of graphattention autoencoders, multilayer perceptron (MLP) classifiers, and saliency map computation by the backpropagated gradient. We illustrate the effectiveness of STAMarker and compare it with three competing methods on four spatial transcriptomic data generated by various platforms. STAMarker considers all genes at once and is more robust when the dataset is very sparse. STAMarker could identify spatial domain-specific SVGs for characterizing spatial domains and enable in-depth analysis of the region of interest in the tissue section.

bioinformatics↗

bioRxiv2022

Prediction of elimination of intrahepatic cccDNA in hepatitis B virus-infected patients by a combination of noninvasive viral markers

Evaluation of intrahepatic covalently closed circular DNA (cccDNA) is a key for searching an elimination of hepatitis B virus (HBV) infection. HBV RNA and HBV core-related antigen have been proposed as surrogate markers for evaluating cccDNA activity, although they do not necessarily estimate the amount of cccDNA. Here, we developed a novel multiscale mathematical model describing intra- and inter-cellular viral propagation, based on the experimental quantification data in both HBV-infected cell culture and humanized mouse models. We applied it to HBV-infected patients under treatment and developed a model which can predict intracellular HBV dynamics only by use of noninvasive extracellular surrogate biomarkers. Importantly, the model prediction of the amount of cccDNA in patients over time was confirmed to be well-correlated with the liver biopsy data. Thus, our noninvasive method enables to predict the amount of cccDNA in patients and contributes to determining the treatment endpoint required for elimination of intrahepatic cccDNA.

microbiology↗

bioRxiv2021

Early Dynamics of Chronic Myeloid Leukemia on Nilotinib Predicts Deep Molecular Response

Chronic myeloid leukemia (CML) is a myeloproliferative disorder caused by the BCR-ABL1 tyrosine kinase1,2. ABL1-selective tyrosine kinase inhibitors (TKIs) including nilotinib have dramatically improved the prognosis of patients with CML3-7. The ultimate goal of CML treatment is likely to be TKI-free maintenance of deep molecular response (DMR), which is defined by quantitative measurement of BCR-ABL1 transcripts on the international scale (IS)8, and durable DMR is a prerequisite to reach this goal9. Thus, an algorithm to enable the early prediction of DMR achievement on TKI therapy is quite valuable. Here, we show that our mathematical framework based on a clinical trial dataset10 can accurately predict the response to frontline nilotinib. We found that our simple dynamical model can predict nilotinib response by using two common laboratory findings (observation values): IS11,12 and white blood cell (WBC) count. Furthermore, our proposed method identified patients who failed to achieve DMR with high fidelity according to the data collected only at three initial time points during nilotinib therapy. Since our model relies on the general properties of TKI response, our framework would be applicable to CML patients who receive frontline nilotinib or other TKIs in clinical practice. Significance StatementChronic myeloid leukemia (CML) is a myeloproliferative disorder caused by the BCR-ABL1 tyrosine kinase. The goal of this treatment is the sequential achievement of deep molecular response (DMR). Tyrosine kinase inhibitors (TKIs) are effective in the reduction because they inhibit CML cell proliferation. However, because of individual differences in the TKI efficacy, some patients are unable to achieve DMR, so that early prediction of DMR reachability is necessary for personalized medicine. By combining time series analysis and mathematical modeling, we developed a mathematical method that accurately predicts patients who do not achieve DMR in the early stages of TKI administration. Our prediction method gives a basis of effective personalized treatments for CML patients.

systems biology↗

bioRxiv2020

Multidrug treatment with nelfinavir and cepharanthine against COVID-19

Antiviral treatments targeting the emerging coronavirus disease 2019 (COVID-19) are urgently required. We screened a panel of already-approved drugs in a cell culture model of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and identified two new antiviral agents: the HIV protease inhibitor Nelfinavir and the anti-inflammatory drug Cepharanthine. In silico modeling shows Nelfinavir binds the SARS-CoV-2 main protease consistent with its inhibition of viral replication, whilst Cepharanthine inhibits viral attachment and entry into cells. Consistent with their different modes of action, in vitro assays highlight a synergistic effect of this combined treatment to limit SARS-CoV-2 proliferation. Mathematical modeling in vitro antiviral activity coupled with the known pharmacokinetics for these drugs predicts that Nelfinavir will facilitate viral clearance. Combining Nelfinavir/Cepharanthine enhanced their predicted efficacy to control viral proliferation, to ameliorate both the progression of disease and risk of transmission. In summary, this study identifies a new multidrug combination treatment for COVID-19.

microbiology↗

bioRxiv2019

Electrical coupling controls dimensionality and chaotic firing of inferior olive neurons

One of the main challenges facing online neural learning systems with numerous modifiable parameters (or "degrees-of-freedom") such as the cerebellum, is how to avoid "overfitting". We previously proposed that the cerebellum controls the degree-of-freedoms during learning by gradually modulating the electric coupling strength between inferior olive neurons. Here, we develop a modeling technique to estimate effective coupling strengths between inferior olive neurons from in vivo recordings of Purkinje cell complex spike activity in three different coupling conditions. We show that high coupling strengths induce synchronous firing and decrease the dimensionality of inferior olive firing dynamics. In contrast, intermediate coupling strengths induce chaotic firing and increase the dimensionality of firing dynamics. Our results thus support the hypothesis that effective coupling controls the dimensionality of inferior olive firing, which may allow the olivocerebellar system to learn effectively from a small training sample set despite the low firing frequency of inferior olive neurons.

neuroscience↗

bioRxiv2017

Context-Dependent Attractor Dynamics in Visual Cortex

The capacity for flexible sensory-action association in animals has been related to context-dependent attractor dynamics outside the sensory cortices. Here we report a line of evidence that flexibly modulated attractor dynamics during task switching are already present in the higher visual cortex in macaque monkeys. With a nonlinear decoding approach, we can extract the particular aspect of the neural population response that reflects the task-induced emergence of bistable attractor dynamics in a neural population, which could be obscured by standard unsupervised dimensionality reductions such as PCA. The dynamical modulation selectively increases the information relevant to task demands, indicating that such modulation is beneficial for perceptual decisions. A computational model that features nonlinear recurrent interaction among neurons with a task-dependent background input replicates the key properties observed in the experimental data. These results suggest that the context-dependent attractor dynamics involving the sensory cortex can underlie flexible perceptual abilities.

neuroscience↗