Biology subjects

Kojima, Y.

Publications and source records attributed to Kojima, Y..

At least 19 recordsLinked to original sources

bioRxiv2026

Probabilistic coupling of cellular and microenvironmental heterogeneity by masked self-supervised learning

AO_SCPLOWBSTRACTC_SCPLOWSpatial omics technologies have advanced to single-cell resolution, enabling systematic analysis of tissue microenvironments alongside cellular-state heterogeneity. However, computationally defining microenvironmental states at single-cell resolution and identifying representations most informative for biological discovery remain major challenges. Here we present Mievformer, a Transformer-based masked self-supervised framework that learns microenvironmental embeddings by encoding neighboring cellular states and relative spatial configurations to parameterize the conditional distribution of continuous cell states at central spatial positions. Through InfoNCE optimization, Mievformer learns representations that capture the relative enrichment of cell states across microenvironments, formalized as a conditional density ratio, thereby enabling probabilistic inference of the coupling between microenvironmental and cellular heterogeneity. Mievformer outperformed existing methods in niche clustering on simulated spatial transcriptomics data and achieved the highest average performance across five real datasets spanning three spatial transcriptomics platforms when evaluated using DREC, a ground-truth-free metric that most strongly correlated with ground-truth performance in simulations. Beyond conventional clustering, Mievformer enables identification of cellular subpopulations based on their microenvironmental distribution and detection of gene-expression signatures associated with colocalization of specific cell populations. Together, these results establish Mievformer as a quantitatively robust and biologically informative framework for learning microenvironment representations in spatial omics.

bioinformatics↗

bioRxiv2026

Hierarchical processing and polarization encoding in the cephalopod visual system

Coleoid cephalopods (octopus, cuttlefish and squid, hereafter cephalopods) have evolved a range of complex visually-guided behaviours, from dexterous hunting to skin-pattern based camouflage and communication1. They have also evolved sensitivity to the polarization of light2, an adaptation thought to help detect camouflaged or semitransparent predators and prey in low visibility underwater environments3-10. How visual information is processed by the cephalopod brain to support their behaviours remains unclear. Here, studying the bigfin reef squid Sepioteuthis lessoniana, we performed calcium imaging and electrophysiological recordings from populations of neurons in the large visual center of the cephalopod brain, the optic lobe (OL). We revealed that the retina-recipient superficial OL contains a diversity of functionally distinct cell types, spatially organized into sub-layers, processing spatio-temporal features of light intensity and possessing polarization angle specificity. More complex features, e.g. direction selectivity, are seen in deeper regions of the OL cortex, which also exhibits spontaneous waves of neural activity in the absence of visual input. Neurons in the downstream OL medulla exhibit visual receptive field sizes and spontaneous activity levels which increase with brain depth, consistent with the hierarchical processing of visual information through the medullas tree-like anatomical organization. Medulla neurons exhibit sensitivity to local decreases in the degree of linear polarization (DoLP), which they integrate additively with light intensity information. Underwater imaging in the squids habitat off the coast of Okinawa, Japan, demonstrate that polarization sensitivity confers a robust short-range boost in object-background contrast over a range of objects and environmental conditions. These findings reveal convergent principles of hierarchical visual processing shared between cephalopods and vertebrates, and highlight how cephalopods utilise their distinct adaptation of polarization sensitivity to solve universal visual challenges underwater.

neuroscience↗

bioRxiv2025

STODE: A Deep Generative Framework for Continuous Spatiotemporal Dynamics in Spatial Transcriptomics

MotivationSpatial transcriptomics offers detailed molecular insights but struggles to capture continuous tissue dynamics due to experimental constraints and destructive sampling. Traditional optimal transport (OT) methods attempt to link time points but often rely on unimodal data and oversimplify complex developmental processes. We propose a spatiotemporal deep generative model that captures the continuous evolution of tissue, integrating both spatial and gene-expression dynamics. We first embed spatial transcriptomic particles into a latent space using a VAE and then learn time-dependent dynamics with a potential-guided neural ODE. This framework reveals diverse tissue behaviors, including growth patterns, spatial reorganization, and temporal gene-expression changes. A neural decoder estimates key behaviors like growth, disappearance, movement, and clustering, while a differentiable growth module models region appearance and disappearance. ResultsSTODE is a deep generative model that reconstructs continuous spatiotemporal dynamics from discrete spatial transcriptomics measurements collected at multiple time points. On a complex synthetic dataset simulating gene-driven tissue morphogenesis and mouse midbrain development dataset, STODE significantly outperformed both linear interpolation and Entropic Optimal Transport models in predicting gene expression and spatial coordinates for a excluded timepoint. When applied to a mouse organogenesis dataset, the models backward simulations from late-stage embryos to early progenitors accurately reconstructed known developmental processes. The simulations correctly identified early pharyngeal arch derivatives as the populations occupying the largest area and captured the progressive differentiation of diverse cell types from multipotent neural crest and mesenchymal progenitors, demonstrating the models capability to reconstruct continuous and meaningful biological trajectories from discrete-timepoint spatial transcriptomes. In addition, the inferred factors reflected coordinated activation of morphogenetic programs, including cytoskeletal remodeling and cell-matrix interactions, and revealed transcriptional signatures associated with epithelial-mesenchymal transition and directed tissue migration, suggesting that the model effectively disentangled regulatory modules underlying the spatial organization and morphogenesis of craniofacial structures. The proposed model, STODE, will be broadly applicable to studies on embryogenesis, tissue regeneration, and disease progression, where reconstructing continuous spatiotemporal trajectories is essential for understanding biological mechanisms. Availability and ImplementationThe STODE model was implemented in Python using the PyTorch deep learning library. This code is available on GitHub at https://github.com/LzrRacer/STODE/. Contactk.majima1214@gmail.com, yakojim@ncc.go.jp, shimamura.csb@tmd.ac.jp

bioinformatics↗

bioRxiv2024

mmVelo: A deep generative model for estimating cell state-dependent dynamics across multiple modalities

Single-cell multiomics provides unique insight into the regulatory relationships across different biological layers such as the transcriptome and regulome. However, single-cell multiomics is limited by its ability to capture only static snapshots at the time of observation, restricting the reflection of dynamic state changes orchestrated across modalities. RNA velocity analysis of single cells allows for the prediction of temporal changes in the transcriptome; however, the inferred dynamics cannot be applied across all biological layers, specifically in the regulome. Therefore, to address this limitation, we developed multimodal velocity of single cells (mmVelo), a deep generative model designed to estimate cell state-dependent dynamics across multiple modalities. mmVelo estimates cell state dynamics based on spliced and unspliced mRNA expression, and uses multimodal representation learning to project these dynamics onto chromatin accessibility, inferring chromatin velocity at a single-peak resolution. We applied mmVelo to single-cell multiomics data from a developing mouse brain and validated the accuracy of the estimated chromatin accessibility dynamics. Furthermore, using the estimated dynamics, we identified the transcription factors that are crucial for chromatin accessibility regulation in mouse skin. Finally, using multiomics data as a bridge, we demonstrated that during human brain development, the dynamics of missing modalities can be inferred from single-modal data via cross-modal generation. Overall, mmVelo enhances our understanding of the dynamic interactions between modalities, offering insights into the regulatory relationships across molecular layers.

bioinformatics↗

bioRxiv2024

scSurv: a deep generative model for single-cell survival analysis

Single-cell omics analysis has unveiled the heterogeneity of various cell types within tumors. However, no methodology currently reveals how this heterogeneity influences cancer patient survival at single-cell resolution. Here, we introduce scSurv, combining a Cox proportional hazards model with a deep generative model of single-cell transcriptome, to estimate individual cellular contributions to clinical outcomes. The accuracy of scSurv was validated using both simulated and real datasets. This method identifies cells associated with favorable or adverse prognoses and extracts genes correlated with their contribution levels. In melanoma, scSurv reproduces known prognostic macrophage classifications and facilitates hazard mapping through spatial transcriptomics in renal cell carcinoma. We also identified genes consistently associated with prognosis across multiple cancers and demonstrated the applicability of this method to infectious diseases. scSurv is a novel framework for quantifying the heterogeneity of individual cellular effects on clinical outcomes.

bioinformatics↗

bioRxiv2024

BST2 induces vascular smooth muscle cell plasticity and phenotype switching during cancer progression

BackgroundSmooth muscle cell (SMC) plasticity and phenotypic switching play prominent roles in the pathogenesis of multiple diseases, but their role in tumorigenesis is unknown. We investigated whether and how SMC diversity and plasticity plays a role in tumor angiogenesis and the tumor microenvironment. Methods and ResultsWe use SMC-specific lineage-tracing mouse models and single cell RNA sequencing to observe the phenotypic diversity of SMCs participating in tumor vascularization. We find that a significant proportion of SMCs adopt a phenotype traditionally associated with macrophage-like cells. These cells are transcriptionally similar to resolution phase M2b macrophages, which have been described to have a role in inflammation resolution. Computationally predicted by the ligand-receptor algorithm CellChat, signaling from BST2 on the surface of tumor cells to PIRA2 on SMCs promote this phenotypic transition; in vitro SMC assays demonstrate upregulation of macrophage transcriptional programs, and increased proliferation, migration, and phagocytic ability when exposed to BST2. Knockdown of BST2 in the tumor significantly decreases the transition towards a macrophage-like phenotype, and cells that do transition have a comparatively higher inflammatory signal typically associated with anti-tumor effect. ConclusionAs BST2 is known to be a poor prognostic marker in multiple cancers where it is associated with an M2 macrophage-skewed TME, these studies suggest that phenotypically switched SMCs may have a previously unidentified role in this immunosuppressive milieu. Further translational work is needed to understand how this phenotypic switch could influence the response to anti-cancer agents and if targeted inhibition of SMC plasticity would be therapeutically beneficial.

cancer biology↗

bioRxiv2024

Development of shell field populations in gastropods

The embryonic shell field of mollusks first appears during gastrulation of the dorsal ectoderm and subsequently develops into the shell-secreting mantle in adult animals. Although several lines of evidence have revealed that this shell field lineage is exclusively derived from the second quartet (2q) of the 16-cell embryos, it is generally believed that the establishment of the shell field fate would be accomplished only after receiving inductive signals from the invaginated endoderm. Despite being accepted as a comprehensive model for molluskan shell field specification, the validity of this induction hypothesis remains questionable owing to the lack of clear experimental evidence and contradictory results. Here, we attempted to re-investigate the inductive role of the endoderm in shell field fate establishment in the limpet Nipponacmea fuscoviridis by experimentally disrupting cell-cell contacts between cell lineages after the 16-cell stage. First, we characterized the shell field cell population by performing two-color in situ hybridization. We characterized at least three cell populations in the developing shell field. Using single-cell transcriptome analysis, we identified several specific effector genes for each population, as well as transcription factor genes. Differentiation of each shell field population was inspected in 2q blastomeres isolated from other cells of the 16-cell embryos. Despite the absence of any interlineage interactions (including ectoderm-endoderm contacts), the expression of marker genes for each shell field population was observed in the isolated 2q fragments. In addition, the expression of several shell field genes was detected in embryos in which cytokinesis was blocked at the 16-cell stage. We concluded that the early process of shell field differentiation in the 2q lineage occurs mostly independently of the interactions with other lineages.

developmental biology↗

bioRxiv2024

A deep generative model integrating single-cell time-frequency characteristics transformed from electrophysiological data with transcriptomic features

Patch-seq yields multi-modal data (e.g., transcriptomic and electrophysiological data) from single cells. However, currently used analytical methods are based on a few global electrophysiological features predefined from chronological potential changes and overlook the importance of time-frequency domain. In this study, we present LincSpectr, a deep neural network model that integrates transcriptomic and electrophysiological features with the latent variables of various variational autoencoders. This model combines the two modalities according to the connection between the latent variables of different modalities calculated by attention-like mechanisms and achieves cross-modal predictions and an inverse analysis. We discovered that the predicted electrophysiological features changed continuously along with their transcriptional profiles and that the neighborhood relationships between the latent states of the transcriptional profiles were consistent with those of the electrophysiological features. Inverse analysis of our model enabled the extraction of gene sets affecting specific time-frequency domains; some genes were likely to be involved in neural activity. Our approach represents a potential avenue to facilitate the discovery of molecular mechanisms underlying time-frequency activities in various cell types, enhancing our understanding of their roles in neural function.

bioinformatics↗

bioRxiv2024

Inferring extrinsic factor-dependent single-cell transcriptome dynamics using a deep generative model

1RNA velocity estimation helps elucidate temporal changes in the single-cell transcriptome. However, current methodologies for inferring single-cell transcriptome dynamics ignore extrinsic factors, such as experimental conditions and neighboring cell. Here, we propose ExDyn--a deep generative model integrated with splicing kinetics for estimating cell state dynamics dependent on extrinsic factors. ExDyn enables the counterfactual inference of cell state dynamics under different conditions. Among the extrinsic factors, ExDyn can extract key features which have large effects on cell state dynamics. ExDyn correctly estimated the difference in dynamics between two conditions and showed better accuracy over existing RNA velocity methods. ExDyn were utilized for unveiling the effect of PERK-knockout on neurosphere differentiation, hematopoietic stem cell differentiation driven by chromatin activity and the dynamics of squamous cell carcinoma cells dependent on colocalized neighboring cells. These results demonstrated that ExDyn is useful for analyzing key features in the dynamic generation of heterogeneous cell populations.

bioinformatics↗

bioRxiv2024

LineageVAE: Reconstructing Historical Cell States and Transcriptomes toward Unobserved Progenitors

AO_SCPLOWBSTRACTC_SCPLOWSingle-cell RNA sequencing (scRNA-seq) enables comprehensive characterization of the cell state. However, its destructive nature prohibits measuring gene expression changes during dynamic processes such as embryogenesis. Although recent studies integrating scRNA-seq with lineage tracing have provided clonal insights between progenitor and mature cells, challenges remain. Because of their experimental nature, observations are sparse, and cells observed in the early state are not the exact progenitors of cells observed at later time points. To overcome these limitations, we developed LineageVAE, a novel computational methodology that utilizes deep learning based on the property that cells sharing barcodes have identical progenitors. This approach transforms scRNA-seq observations with an identical lineage barcode into sequential trajectories toward a common progenitor in a latent cell state space. Using hematopoiesis and reprogrammed fibroblast datasets, we demonstrate the capability of LineageVAE to reconstruct unobservable cell state transitions, historical transcriptome, and regulatory dynamics toward progenitor cell states at single-cell resolution.

bioinformatics↗

bioRxiv2023

A deep generative model for estimating single-cell RNA splicing and degradation rates

AO_SCPLOWBSTRACTC_SCPLOWMessenger RNA splicing and degradation are critical for gene expression regulation, the abnormality of which leads to diseases. Previous methods for estimating kinetic rates have limitations, assuming uniform rates across cells. We introduce DeepKINET, a deep generative model that estimates splicing and degradation rates at single-cell resolution from scRNA-seq data. DeepKINET outperformed existing methods on simulated and metabolic labeling datasets. Applied to forebrain and breast cancer data, it identified RNA-binding proteins responsible for kinetic rate diversity. DeepKINET also analyzed the effects of splicing factor mutations on target genes in erythroid lineage cells. DeepKINET effectively reveals cellular heterogeneity in post-transcriptional regulation.

bioinformatics↗

bioRxiv2023

scEGOT: Single-cell trajectory inference framework based on entropic Gaussian mixture optimal transport

Time-series scRNA-seq data have opened a door to elucidate cell differentiation, and in this context, the optimal transport theory has been attracting much attention. However, there remain critical issues in interpretability and computational cost. We present scEGOT, a comprehensive framework for single-cell trajectory inference, as a generative model with high interpretability and low computational cost. Applied to the human primordial germ cell-like cell (PGCLC) induction system, scEGOT identified the PGCLC progenitor population and bifurcation time of segregation. Our analysis shows TFAP2A is insufficient for identifying PGCLC progenitors, requiring NKX1-2. Additionally, MESP1 and GATA6 are also crucial for PGCLC/somatic cell segregation. These findings shed light on the mechanism that segregates PGCLC from somatic lineages. Notably, not limited to scRNA-seq, scEGOTs versatility can extend to general single-cell data like scATAC-seq, and hence has the potential to revolutionize our understanding of such datasets and, thereby also, developmental biology.

bioinformatics↗

bioRxiv2023

Inference of gene regulatory networks using time-series single-cell RNA-seq data with CRISPR perturbations

Single-cell RNA-seq analysis coupled with CRISPR-based perturbation (scCRISPR) has enabled the inference of gene regulatory networks (GRNs) with causal relationships. However, a snapshot of scCRISPR data may not lead to an accurate inference, since a gene knockout can influence multi-layered downstream over time. Here, we developed RENGE, a new computational method that infers GRNs using a time-series scCRISPR dataset. RENGE models the propagation process of the effects elicited by a gene knockout on its GRN. It can distinguish between direct and indirect regulations, which allows for the inference of regulations by genes that are not knocked out. RENGE therefore outperforms current methods in the accuracy of inferring GRNs. When used on a dataset we derived from human-induced pluripotent stem cells, RENGE yielded a GRN consistent with multiple databases and literature. Accurate inference of GRNs by RENGE would enable the identification of key factors for various biological systems.

bioinformatics↗

bioRxiv2022

Extensive gut virome variation and its associations with host and environmental factors in a population-level cohort

Indigenous bacteriophage communities (virome) in the human gut have a huge impact on the gut bacterial communities (bacteriome), but virome variation at a population scale is not fully investigated yet. Here, we analyse the gut virome in a population-level cohort of 4,198 deeply phenotyped individuals. We discovered thousands of high-quality phage genomes including previously uncharacterized ones with different bacterial hosts than known major phage clades. The distribution of host bacteria was a strong determinant for the distribution of phages, and virome diversity was highly correlated with anti-viral defence mechanisms of the bacteriome, such as CRISPR-Cas and restriction-modification systems. We identified 97 various intrinsic/extrinsic factors that significantly affect the virome, including age, sex, lifestyle, and diet, most of which showed consistent associations with both phages and their predicted bacterial hosts. Among the metadata categories, disease and medication had the strongest effects on the virome structure. Overall, these results present a basis to understand the symbiotic communities of bacteria and their viruses in the human gut, which will facilitate the medical and industrial applications of indigenous viruses.

microbiology↗

bioRxiv2022

Resolution of the curse of dimensionality in single-cell RNA sequencing data analysis

Single-cell RNA sequencing (scRNA-seq) can determine gene expression in numerous individual cells simultaneously, promoting progress in the biomedical sciences. However, scRNA-seq data are high-dimensional with substantial technical noise, including dropouts. During analysis of scRNA-seq data, such noise engenders a statistical problem known as the curse of dimensionality (COD). Based on high-dimensional statistics, we herein formulate a noise reduction method, RECODE (resolution of the curse of dimensionality), for high-dimensional data with random sampling noise. We show that RECODE consistently eliminates COD in relevant scRNA-seq data with unique molecular identifiers. RECODE does not involve dimension reduction and recovers expression values for all genes, including lowly expressed genes, realizing precise delineation of cell-fate transitions and identification of rare cells with all gene information. Compared to other representative imputation methods, RECODE employs different principles and exhibits superior overall performance in cell-clustering and single-cell level analysis. The RECODE algorithm is parameter-free, data-driven, deterministic, and high-speed, and notably, its applicability can be predicted based on the variance normalization performance. We propose RECODE as a general strategy for preprocessing noisy high-dimensional data.

bioinformatics↗

bioRxiv2022

Single-cell colocalization analysis using a deep generative model

1Analyzing colocalization of single cells with heterogeneous molecular phenotypes is essential for understanding cell-cell interactions, cellular responses to external stimuli, and their biological functions in diseases and tissues. However, high-throughput methods for identifying spatial proximity at single-cell resolution are practically unavailable. Here, we introduce DeepCOLOR, a computational framework based on a deep generative model that recovers inter-cellular colocalization networks with single cell resolution by the integration of single cell and spatial transcriptomes. It segregates cell populations defined by the colocalization relationships and predicts cell-cell interactions between colocalized single cells. DeepCOLOR could identify plausible cell-cell interaction candidates in mouse brain tissues, human squamous cell carcinoma samples, and human lung tissues infected with SARS-CoV-2 by reconstructing spatial colocalization maps at single-cell resolution. DeepCOLOR is typically applicable to studying cell-cell interactions in any spatial niche. Our newly developed computational framework could help uncover molecular pathways across single cells connected with colocalization networks.

bioinformatics↗

bioRxiv2021

Strip1 regulates retinal ganglion cell survival by suppressing Jun-mediated apoptosis to promote retinal neural circuit formation

In the vertebrate retina, an interplay between retinal ganglion cells (RGCs), amacrine, and bipolar cells establishes a synaptic layer called the inner plexiform layer (IPL). This circuit conveys signals from photoreceptors to visual centers in the brain. However, the molecular mechanisms involved in its development remain poorly understood. Striatin-interacting protein 1 (Strip1) is a core component of the STRIPAK complex, and it has shown emerging roles in embryonic morphogenesis. Here, we uncover the importance of Strip1 in inner retina development. Using zebrafish, we show that loss of Strip1 causes defects in IPL formation. In strip1 mutants, RGCs undergo dramatic cell death shortly after birth. Amacrine and bipolar cells subsequently invade the degenerating RGC layer, leading to a disorganized IPL. Mechanistically, zebrafish Strip1 interacts with its STRIPAK partner, Striatin3, and both show overlapping functions in RGC survival. Furthermore, loss of Strip1 or Striatin 3, leads to activation of the pro-apoptotic marker, Jun, within RGCs, and Jun knockdown rescues RGC survival in strip1 mutants. In addition to its function in RGC maintenance, Strip1 is required for RGC dendritic patterning, which likely contributes to proper IPL formation. Taken together, we propose that a series of Strip1-mediated regulatory events coordinates inner retinal circuit formation by maintaining RGCs during development, which ensures proper positioning and neurite patterning of inner retinal neurons.

developmental biology↗

bioRxiv2021

Ergodic patterns of cell state transitions underlie the reproducibility of embryonic development

The reproducibility of embryonic development is a remarkable feat of biological organization, but the underlying mechanisms are poorly understood. Clearly, gene regulatory networks are central to the orderly progression of development, but noisy molecular and cellular processes should reduce reproducibility. Here, we identify ergodicity, a type of dynamical stability, as underlying the reproducibility of development. In ergodic systems, a single timepoint measurement equals a time average. Focusing on the zebrafish tailbud, we define gene expression and cell motion states using a parallel statistical analyses of single cell RNA sequencing data and in vivo timelapse cell tracking data and a change point detection algorithm. Strikingly, the cell motion state transitions in each embryo exhibit the same patterns for both a single timepoint and a 2-3 hour time average. Both the cell motion and gene expression cell states exhibit balanced influx and outflux rates reflecting a spatiotemporal stability. Stated simply, these data indicate the pattern of changes in the tailbud doesnt change. This ergodic pattern of cell state transitions may represent an emergent meta-state that links gene networks to the reproducible progression of embryogenesis.

developmental biology↗