Biology subjects

Kimura, K. D.

Publications and source records attributed to Kimura, K. D..

8 recordsLinked to original sources

bioRxiv2026

DiverScan: attention-guided exploratory analysis of animal behaviour across diverse comparative experiments

High-throughput tracking yields rich time-series data of animal behaviour, yet extracting biological insight remains a bottleneck, even for simple control--treatment comparisons. Existing analyses struggle to pinpoint the multiple, nonlinearly intertwined differences hidden in the data and fail to generalise across diverse recording set-ups. Here we present DiverScan, a deep-learning framework for attention-guided exploratory analysis of behavioural time series. DiverScan disentangles condition-dependent behavioural differences by pairing a loss function tailored to data exploration with an attention-branch architecture, and highlights condition-specific behavioural features visually and with explanatory text generated by a large language model. DiverScan is validated on a synthetic benchmark and on datasets from seven laboratories spanning multiple species and both single- and multi-animal behaviour. Beyond recovering known phenotypes, we demonstrate that DiverScan helps researchers generate new hypotheses across these diverse datasets. Available as Jupyter notebooks with a graphical user interface and flexible input formats, DiverScan can integrate into existing behavioural-analysis pipelines.

animal behavior and cognition↗

bioRxiv2025

Cellular Responses to Photothermal Therapy: Heat-Induced ERK Signaling and Intercellular Communication in Solid Tumors

Nanoparticle-mediated photothermal therapy (PTT) shows promise as a standalone cancer treatment but faces clinical challenges due to inconsistent efficacy. Its translation is further hindered by a limited understanding of plasmon-induced heat effects, such as stress responses and intercellular signaling. Here, we investigate how plasmon-induced local heating affects cellular behavior and fate within tumor spheroids by focusing on the activity of the extracellular signal-regulated kinase (ERK). Spheroids were prepared from HeLa cells that express a FRET-based ERK sensor, and ERK activity changes under photothermal stimuli were tracked using a deep-learning program, 3DeeCellTracker. Gold nanostars were used as highly efficient photothermal transducers. Our results revealed significant alterations in ERK signaling patterns upon photothermal stimulation compared to spontaneous ERK activity in untreated spheroids, including changes in the activation frequency, timing, and duration. Notably, photothermal-induced ERK activity propagated across neighboring cells within the spheroid, suggesting intercellular communication. Furthermore, analysis of cell death and division further demonstrated that laser power modulates cellular fate during photothermal therapy. This study provides insights for predicting the therapeutic effects of PPT and guides the rational design of next-generation photothermal strategies. Additionally, our approach demonstrates the potential of FRET-based biosensors and deep-learning tools as powerful methods to study the effects of various therapeutic stimuli on solid tumors at the single-cell level.

cancer biology↗

bioRxiv2022

Seg2Link: an efficient and versatile solution for semi-automatic cell segmentation in 3D image stacks

Recent advances in microscopy techniques, especially in electron microscopy, are transforming biomedical studies by acquiring large quantities of high-precision 3D cell image stacks. However, to study cell morphology and connectivity in organs such as brains, scientists must first perform cell segmentation, which involves extracting individual cell regions of various shapes and sizes from a 3D image. This remains a great challenge because automatic cell segmentation can contain numerous errors, even with advanced deep learning methods. For biomedical research that requires cell segmentation in large 3D image stacks, an efficient semi-automated software solution is still needed. We created Seg2Link, which generates automatic segmentations based on deep learning predictions and allows users to quickly correct errors in the segmentation results. It can perform automatic instance segmentation of 2D cells in each slice, 3D cell linking across slices, and various manual corrections, in order to efficiently transform inaccurate deep learning predictions into accurate segmentation results. Seg2Links data structure and algorithms were also optimized to process 3D images with billions of voxels on a personal computer quickly. Thus, Seg2Link offers a simple and effective way for scientists to study cell morphology and connectivity in 3D image stacks.

neuroscience↗

bioRxiv2022

Optogenetic control of cell signaling with red/far-red light-responsive optogenetic tools in Caenorhabditis elegans.

Optogenetic techniques have been intensively applied to the nematode Caenorhabditis elegans to investigate its neural functions. However, as most of these optogenetics are responsive to blue light and the animals exhibits avoidance behavior to blue light, the application of optogenetic tools responsive to longer wavelength light has been eagerly anticipated. In this study, we report the implementation in C. elegans of a phytochrome-based optogenetic tool that responds to red/near-infrared light and manipulates cell signaling. We first introduced the SynPCB system, which enabled us to synthesize phycocyanobilin (PCB), a chromophore for phytochrome, and confirmed the biosynthesis of PCB in neurons, muscles, and intestinal cells. We further confirmed that the amount of PCBs synthesized by the SynPCB system was sufficient for photoswitching of phytochrome B (PhyB)-phytochrome interacting factor 3 (PIF3). In addition, optogenetic elevation of intracellular Ca2+ levels in intestinal cells induced a defecation motor program. These SynPCB system and phytochrome-based optogenetic techniques would be of great value in elucidating the molecular mechanisms underlying C. elegans behaviors.

synthetic biology↗

bioRxiv2021

Neural mechanism of experience-dependent sensory gain control in C. elegans

Animals sensory systems adjust their responsiveness to environmental stimuli that vary greatly in their intensity. Here we report the neural mechanism of experience-dependent sensory adjustment, especially gain control, in the ASH nociceptive neurons in Caenorhabditis elegans. Using calcium imaging under gradual changes in stimulus intensity, we find that the ASH neurons of naive animals respond to concentration increases in a repulsive odor 2-nonanone regardless of the magnitude of the concentration increase. However, after preexposure to the odor, the ASH neurons exhibit significantly weak responses to a small gradual increase in odor concentration while their responses to a large gradual increase remain strong. Thus, preexposure changes the slope of stimulus-response relationships (i.e., gain control). Behavioral analysis suggests that this gain control contributes to the preexposure-dependent enhancement of odor avoidance behavior. Mathematical analysis reveals that the ASH response consists of fast and slow components, and that the fast component is specifically suppressed by preexposure. In addition, genetic analysis suggests that G protein signaling is required for the fast component. Thus, our integrative study demonstrates how prior experience dynamically modulates stimulus-response relationships in sensory neurons, eventually leading to adaptive modulation of behavior.

neuroscience↗

bioRxiv2021

Electric shock causes fear-like persistent behavioral response in the nematode Caenorhabditis elegans

Behavioral persistency reflects internal brain states, which are the foundations of multiple brain functions. However, experimental paradigms that enable genetic analyses of behavioral persistency and its associated brain functions have been limited. Here we report novel persistent behavioral responses caused by electric stimuli in the nematode Caenorhabditis elegans. When the animals on bacterial food are stimulated by alternating current, their movement speed suddenly increases more than 2-fold, which persists for minutes even after the electric stimulation is terminated. Genetic analyses reveal that multiple types of voltage-gated channels are required for the response, possibly as the sensors, and neuropeptide signaling regulates the duration of the persistent response. Additional behavioral analyses indicate that the animals response to electric shock is scalable and has a negative valence. These properties, along with persistence, have been recently regarded as essential features of emotion, suggesting that the animals response to electric shock may express a form of emotion, such as fear.

neuroscience↗

bioRxiv2018

Deep-learning-based flexible pipeline for segmenting and tracking cells in 3D image time series for whole brain imaging

The brain is a complex system that operates based on coordinated neuronal activities. Brain-wide cellular calcium imaging techniques have quickly advanced in recent years and become powerful tools for understanding the neuronal activities of small animal models. The whole brain imaging generally requires to extract the neuronal activities from three-dimensional (3D) image series. Unfortunately, the 3D image series are obtained under imaging conditions different among laboratories and extracting neuronal activities from the data requires multiple processes. Therefore researchers need to develop their own software, which has prevented the application of whole-brain imaging experiments in more laboratories. Here, we combined traditional image processing techniques with the powerful deep-learning method which can be flexibly modified to fit 3D image data in the nematode Caenorhabditis elegans obtained under different conditions. We first trained the 3D U-net deep network to classify each pixel into cell and non-cell categories. Cells merged as a whole region were further separated into individual cells by watershed segmentation. The cells were then tracked in 3D space over time with the combination of a feedforward network and a point set registration method to use local and global relative positions of the cells, respectively. Remarkably, one manually annotated 3D image combined with data augmentation was sufficient for training the deep networks to obtain satisfactory tracking results. Our method correctly tracked more than 98% of neurons in three different image datasets and successfully extracted brain-wide neuronal activities. Our method worked well even when the sampling rate was reduced: 86% correct in case 4/5 frames were removed, and when artificial noise was added into the raw images: 91% correct in case 35 times of background-level noise was added. Our results proved that deep learning is widely applicable to different datasets and can help us in establishing a flexible pipeline for extracting whole brain activities.

bioinformatics↗

bioRxiv2017

Experience-dependent modulation of behavioral features in sensory navigation of nematodes and bats revealed by machine learning

Animal behavior is the final and integrated output of the brain activity. Thus, recording and analyzing behavior is critical to understand the underlying brain function. While recording animal behavior has become easier than ever with the development of compact and inexpensive devices, detailed behavioral data analysis requires sufficient previous knowledge and/or high content data such as video images of animal postures, which makes it difficult for most of the animal behavioral data to be efficiently analyzed to understand brain function. Here, we report a versatile method using a hybrid supervised/unsupervised machine learning approach to efficiently estimate behavioral states and to extract important behavioral features only from low-content animal trajectory data. As proof of principle experiments, we analyzed trajectory data of worms, fruit flies, rats, and bats in the laboratories, and penguins and flying seabirds in the wild, which were recorded with various methods and span a wide range of spatiotemporal scales--from mm to 1000 km in space and from sub-seconds to days in time. We estimated several states during behavior and comprehensively extracted characteristic features from a behavioral state and/or a specific experimental condition. Physiological and genetic experiments in worms revealed that the extracted behavioral features reflected specific neural or gene activities. Thus, our method provides a versatile and unbiased way to extract behavioral features from simple trajectory data to understand brain function.

neuroscience↗