Biology subjects

Koda, H.

Publications and source records attributed to Koda, H..

17 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

Rats spontaneously show categorical responses toward familiar or unfamiliar conspecifics in a habituation-dishabituation task using multiple habituation stimuli

Discriminating between familiar and unfamiliar conspecifics is a fundamental aspect of social recognition. Previous studies have shown that familiarity strongly influences rat social behavior, and that rats can be trained to discriminate between olfactory cues from cage-mates and non-cage-mates in a digging task. However, such discrimination did not occur in a lever-pressing task using live conspecifics as stimuli. In the present study, we investigated whether rats could discriminate social familiarity in a simple, less-demanding and naturalistic context using whole-animal stimuli. We employed a habituation-dishabituation paradigm in which multiple cage-mate or non-cage-mate individuals were presented during the habituation phase. We examined whether rats would habituate to the familiarity category among these multiple stimuli and subsequently show dishabituation when the stimulus class switched (cage-mate to non-cage-mate or vice versa). During the habituation phase, exploratory responses declined over successive sessions, suggesting habituation to the common element among multiple cage-mates or non-cage-mates, i.e., familiarity/unfamiliarity. When all subjects were considered together, clear dishabituation was not observed, indicating individual variation in habituation. However, a post-hoc analysis restricted to subjects showing sufficient habituation revealed significant dishabituation to the stimulus class switch. These findings suggest that rats can spontaneously classify multiple conspecifics according to their familiarity without training. Rats may flexibly employ different forms of social recognition and information such as individual identity or familiarity-based heuristics to guide adaptive social behavior in varying contexts. HighlightsO_LIWe tested familiarity discrimination in rats by habituation-dishabituation task C_LIO_LIMultiple cage-mate or non-cage-mate stimuli were presented during habituation C_LIO_LIExploratory responses decreased as habituation sessions progressed C_LIO_LIDishabituation occurred in individuals who exhibited sufficient habituation C_LIO_LIRats showed categorical responses toward familiarity or unfamiliarity C_LI

animal behavior and cognition↗

bioRxiv2025

Analysis of individual identification and age-class classification of wild macaque vocalizations without pitch- and formant-based acoustic parameter measurements

In recent years, deep learning has achieved high performance in bioacoustic classification tasks by leveraging automatically processed acoustic features for large datasets. However, few performance evaluations of automatically processed acoustic features have been conducted on small-scale data because deep learning requires large datasets. To test whether mel spectrograms (an automatically processed acoustic features) are effective for classifying relatively small acoustic data, we evaluated the performance of two classification machines (random forest and support vector machine) using mel-spectrograms of 651 coo calls of six wild female Japanese macaques on two tasks: 1) individual identification and 2) age-class classification between younger (<10 yrs) and the older animals (>20 yrs). For the individual identification task, the mean balanced accuracy was 0.81 for random forest and 0.82 for support vector machine. For the age-class classification task, the mean balanced accuracy was 0.91 for random forest and 0.93 for support vector machine. Considering that of all the calls were recorded in the wild, methods using automatically processed acoustic features, such as mel spectrogram, are effective in classifying small acoustic data for the individual identification task. The high performance in the age-class classification task might be attributable to the potential of mel spectrograms to capture the characteristics of older individuals (e.g. harshness).

animal behavior and cognition↗

bioRxiv2024

How does immediate auditory feedback coupled to locomotion influence the spontaneous actions of rats?

Sense of agency is a subjective experience that is difficult to examine in animals and preverbal infants. In animals and infants, the sensory monitoring of self-motion, which is the basis of this sense, and the contingency between self-motion and sensory feedback are investigated. Here, we performed experiments on rats using locomotor movements and auditory feedback of pulsed sounds with real-time changes in pulse interval corresponding to the velocity of locomotion, and tested whether rats change their locomotion depending on the degree of contingency between locomotion and auditory feedback. The results showed that rats increased their locomotion distance when the pulse interval of sound changed in accordance with the velocity of locomotion. However, this effect was not long-lasting. This study provides a simple and convenient method to examine the awareness of the contingency between self-motion and auditory feedback in a wide range of animal species without long-term training, and is expected to be a useful tool for comparisons with humans.

animal behavior and cognition↗

bioRxiv2024

Evolution of Key Factors Influencing Performance Across Phases in Junior Short Sprints

Sprint performance plays a crucial role in various sports. Short sprints vary in court size and competitive characteristics but are common in many sports. Although the relationship between age and muscle strength has been explored in short sprints, there is limited understanding of how various physical factors interact, particularly concerning differences in the acceleration phase. This study examined the relationship between sprint times at 0-2.5 m, 2.5-5 m, and 5-10 m intervals and various factors in junior athletes. Results indicated that sprint times increased with age, and is correlated with muscle strength and flexibility. Partial correlation analysis showed that faster times in the 0-2.5 m interval were associated with higher hip flexibility; in the 2.5-5 m interval, faster times were associated with higher core flexibility; and in the 5-10 m interval, a relationship with standing long jump performance was confirmed. Furthermore, lower fat-free body weight translated to higher performance. In the acceleration phase of 10 m, flexibility immediately after the start and the subsequent horizontal propulsive force is important factors that are strongly related to performance change in each interval. The study results illuminate the mechanism behind short sprints in junior athletes and emphasize the importance of daily training, considering the sprint distance required for each sport.

animal behavior and cognition↗

bioRxiv2024

Contagious vocal reaction forms avian vocal exchange: an experimental study in Java sparrows

In human turn-taking, speakers alternate at very short intervals while avoiding overlaps. If speakers do not receive a vocal response from another party, they often repeat their utterance after the expected response time has elapsed to elicit a reply. Intra-individual intervals tend to be longer than inter-individual intervals. Such temporal regularity in vocal exchanges has also been observed in social mammals, such as non-human primates. In contrast, vocal exchanges in birds have been studied primarily in the context of songs, with limited research on call-based vocal exchanges. Studies specifically examining intra-individual call intervals are even scarcer. In this study, we investigated vocal exchanges and their temporal patterns in Java sparrows placed face-to-face. The results revealed that they vocalized at very short intervals following the calls of the other individual. However, no significant differences were observed between the inter- and intra-individual intervals. This contrasts with the temporal characteristics of vocal exchanges observed in social mammals that have been studied to date. We propose that a possible explanation for this difference lies in the variation in social group structures between birds and mammals.

animal behavior and cognition↗

bioRxiv2024

Familiarity modulates preference for joint feeding with conspecifics in rats

Rats are highly social mammals that form social groups of various sizes. Interestingly, they are known to show prosocial behaviors, and are socially tolerant of other conspecifics as they feed together, but the specificity of their prosociality has not been fully understood. Here, we investigated what degree rats showed social choice, i.e., preferential joint feeding even with unfamiliar animals by performing a simple feeding site choice task. In the task, subjects were tested to choose between the feeding site where they fed with other rats (social option) or the other site where they could feed alone (solitary option). The results showed that social option choices increased particularly when the partner was a single non-cagemate rat. This result means that high social preference for other individuals occurs spontaneously, even when they are feeding with unfamiliar non-cagemates. This high degree of social tolerance would lead to the suppression of aggressive behavior as well as enhancement of affiliative relationships even in situations when the rats met with unfamiliar conspecifics for the first time. HighlightsO_LIRats are highly socially tolerant mammals as they feed with conspecifics together. C_LIO_LIWe explored how the presence of conspecifics affects feeding site choices in rats. C_LIO_LISocial choices increased when the partners were non-cagemates. C_LIO_LISocial choices decreased as the number of non-cagemates increased. C_LIO_LIJoint feeding with a single non-cagemate is observed particularly frequently. C_LI

animal behavior and cognition↗

bioRxiv2021

Effects of short-term isolation on social animals' behavior: an experimental case study of Japanese macaque

One of the goals in animal socioecology is to understand the functions and dynamics of group living. While observations of free-ranging animals are a crucial source of information, an experimental investigation that manipulates the size or composition, or both, of animal groups in captivity can also bring complementary contributions to the research inquiry. When paired with an automatic data collection by biologging technology, experimental studies on captive animals also allow for big data analyses based on recent machine learning techniques. As an initial exploration of this research paradigm, the present study inquired to what extent isolation of captive Japanese macaques (Macaca fuscata) changed their movement patterns. Using three-dimensional location trajectories of the macaques that were systematically collected via Bluetooth Low Energy beacons and a deep neural network, we estimated the identifiability of whether a macaque was behaving in isolation or in group. We found that the neural network identified the isolation vs. in-group conditions with more than 90% accuracy from a five-minute location trajectory, suggesting that the isolation caused notable changes from the canonical group-living behaviors. In addition, the isolation made each individual more identifiable from one another based on their location trajectories.

animal behavior and cognition↗

bioRxiv2020

Deep Texture Representations as a Universal Encoder for Pan-cancer Histology

Cancer histological images contain rich biological and clinical information, but quantitative representation can be problematic and has prevented direct comparison and accumulation of large-scale datasets. Here we show that deep texture representations (DTRs) produced by a bilinear Convolutional Neural Network, express cancer morphology well in an unsupervised manner, and work as a universal encoder for cancer histology. DTRs are useful for content-based image retrieval, enabling quick retrieval of histologically similar images from optimally area selected datasets of 7,175 cases from The Cancer Genome Atlas. Via comprehensive comparison with driver and clinically actionable gene mutations, we have successfully predicted 309 combinations of genomic features and cancer types from hematoxylin and eosin-stained images at high accuracy (AUC > 0.70 and q < 0.02). With its mounting capabilities on accessible devices such as smartphones, DTR-based encoding for cancer histology has a potentially strong impact on global equalization for cancer diagnosis and targeted therapies.

cancer biology↗

bioRxiv2020

Measuring Long Context Dependency in Birdsong Using an Artificial Neural Network with a Long-Lasting Working Memory

Context dependency is a key feature in sequential structures of human language, which requires reference between words far apart in the produced sequence. Assessing how long the past context has an effect on the current status provides crucial information to understand the mechanism for complex sequential behaviors. Birdsongs serve as a representative model for studying the context dependency in sequential signals produced by non-human animals, while previous reports were upper-bounded by methodological limitations. Here, we newly estimated the context dependency in birdsongs in a more scalable way using a modern neural-network-based language model whose accessible context length is sufficiently long. The detected context dependency was beyond the order of traditional Markovian models of birdsong, but was consistent with previous experimental investigations. We also studied the relation between the assumed/auto-detected vocabulary size of birdsong (i.e., fine-vs. coarse-grained syllable classifications) and the context dependency. It turned out that the larger vocabulary (or the more fine-grained classification) is assumed, the shorter context dependency is detected.

animal behavior and cognition↗

bioRxiv2020

Agent-based simulation for reconstructing social structure by observing collective movements with special reference to single-file movement

Understanding social organization is fundamental for the analysis of animal societies. In this study, animal single-file movement data -serialized order movements generated by simple bottom-up rules of collective movements-- are informative and effective observations for the reconstruction of animal social structures using agent-based models. For simulation, artificial 2-dimensional spatial distributions were prepared with the simple assumption of clustered structures of a group. Animals in the group are either independent or dependent agents. Independent agents distribute spatially independently each one another, while dependent agents distribute depending on the distribution of independent agents. Artificial agent spatial distributions aim to represent clustered structures of agent locations -a coupling of "core" or "keystone" subjects and "subordinate" or "follower" subjects. Collective movements were simulated following two simple rules, 1) initiators of the movement are randomly chosen, and 2) the next moving agent is always the nearest neighbor of the last moving agents, generating "single-file movement" data. Finally, social networks were visualized, and clustered structures reconstructed using a recent major social network analysis (SNA) algorithm, the Louvain algorithm, for rapid unfolding of communities in large networks. Simulations revealed possible reconstruction of clustered social structures using relatively minor observations of single-file movement, suggesting possible application of single-file movement observations for SNA use in field investigations of wild animals.

animal behavior and cognition↗

bioRxiv2020

Non-Parametric Analysis of Inter-Individual Relations Using an Attention-Based Neural Network

O_LISocial network analysis, which has been widely adopted in animal studies over the past decade, enables the revelation of global characteristic patterns of animal social systems from pairwise inter-individual relations. Animal social networks are typically drawn based on geometric proximity and/or frequency of social behaviors (e.g., grooming), but the appropriate metric for inter-individual relationship is not clear, especially when prior knowledge on the species/data is limited. C_LIO_LIIn this study, researchers explored a non-parametric analysis of inter-individual relations using a neural network with the attention mechanism, which plays a central role in natural language processing. The high interpretability of the attention mechanism and flexibility of the entire neural network allow for automatic detection of inter-individual relations included in the raw data, without requiring prior knowledge/assumptions about what modes/types of relations are included in the data. For these case studies, three-dimensional location data collected from simulated agents and real Japanese macaques were analyzed. C_LIO_LIThe proposed method successfully recovered the latent relations behind the simulated data and discovered female-oriented relations in the real data, which are in accordance with previous generalizations about the macaque social structure. C_LIO_LIThe proposed method does not exploit any behavioral patterns that are particular to Japanese macaques, and researchers can use it for location data of other animals. The exibility of the neural network would also allow for its application to a wide variety of data with interacting components, such as vocal communication. C_LI

animal behavior and cognition↗

bioRxiv2020

Cooperation for copulation: a novel ecological mechanism underlying the evolution of coalition for sharing mating opportunities

In multi-male multi-female groups of animals, male reproductive success is often skewed toward top-ranking males. Coalition formation by middle- to low-rankers can be seen as a collaborative effort to modify the distribution of reproductive success within the group, so that they gain more than they would do otherwise. It has been pointed out, on the other hand, that a coalition of top-ranking males could hardly be profitable in the sense that they would gain little additional benefit from making collaborative effort. Here we report our novel observation of facultative sharing of mating opportunities among males in a wild population of stump-tailed macaques (Macaca arctoides) as a possible case of coalition-like behavior in which dominant males jointly guard females from mating with subordinate males and actively share mating opportunities within the allies. First, we report our novel observation of facultative sharing of mating opportunities in male stump-tailed macaques, where two or more males remain in close proximity to and copulate with a female in turn without contesting or sneaking. Second, considering the kind of coalition formation in which dominant males collaboratively exclude subordinates from mating competition and thereby strengthen the reproductive skew that already exists, we specify, by means of mathematical modeling, the condition for this kind of coalition formation to be selectively favored. Finally, we derive predictions about the occurrence of the coalition-like behavior depending on ecological and demographic factors, and test them empirically using data from the five groups of stump-tailed macaques in our study population.

animal behavior and cognition↗

bioRxiv2019

The nose is mightier than the tooth: larger male proboscis monkeys have smaller canines

The uniquely enlarged noses of male proboscis monkeys are prominent adornments, and a sexually selected male trait. A recent study showed significant correlations among nose, body, and testis sizes and clear associations between nose size and the number of females in a males harem. However, to date, the analyses of other common male traits, i.e., canines, are lacking. Whereas male nose size had a positive correlation with body size, we unexpectedly found a negative correlation between body and canine sizes. We explain this by an interaction between sexual and natural selection. Larger noses in males may interfere with the use of canines, thereby reducing their effectiveness as weapons. Additionally, longer canines are opposed by natural selection because the larger gape it imposes upon its bearer reduces foraging efficiency, particularly in folivores. This unique case of decoupling of body and canine size reveals that large canines carry an ecological cost.

evolutionary biology↗

bioRxiv2019

Animals exhibit personality traits in their movement: a case study on location trajectories of primates

Researching individual recognition (IR) is essential to understand the life history and adaptive behavior of social animals. Investigation of personality traits may also provide insights into how social animals distinguish between different individuals. This study investigates IR behavior in Japanese macaques (Macaca fuscata), focusing on one specific trait, which is movement. Using a recently developed tracking system based on Bluetooth(R) Low Energy beacons, we collected three-dimensional (3D) location data from five Japanese macaques living in a group cage. A non-parametric, neural network-based analysis of the data revealed the existence of personality traits in extremely limited aspects of the movement data (2-min trajectory of 3D location). Our results support the validity of multimodal approaches in studying IR, beyond the typical single-frame face recognition method, both for researchers and animal agents.

animal behavior and cognition↗

bioRxiv2018

Luigi: Large-scale histopathological image retrieval system using deep texture representations

BackgroundAs a large number of digital histopathological images have been accumulated, there is a growing demand of content-based image retrieval (CBIR) in pathology for educational, diagnostic, or research purposes. However, no CBIR systems in digital pathology are publicly available.\n\nResultsWe developed a web application, the Luigi system, which retrieves similar histopathological images from various cancer cases. Using deep texture representations computed with a pre-trained convolutional neural network as an image feature in conjunction with an approximate nearest neighbor search method, the Luigi system provides fast and accurate results for any type of tissue or cell without the need for further training. In addition, users can easily submit query images of an appropriate scale into the Luigi system and view the retrieved results using our smartphone application. The cases stored in the Luigi database are obtained from The Cancer Genome Atlas with rich clinical, pathological, and molecular information. We tested the Luigi system by querying typical cancerous regions from four cancer types, and confirmed successful retrieval of relevant images.\n\nConclusionsThe Luigi system will help students, pathologists, and researchers easily retrieve histopathological images of various cancers similar to those of the query image.

pathology↗

bioRxiv2017

Capturing the Difference in Humoral Immunity between Normal and Tumor Environments from RNA Sequences of B-Cell Receptors Using Supervised Machine Learning

The recent success of immunotherapy in treating tumors has attracted increasing interest in research related to the adaptive immune system in the tumor microenvironment. Recent advances in next-generation sequencing technology enabled the sequencing of whole T-cell receptors (TCRs) and B-cell receptors (BCRs)/immunoglobulins (Igs) in the tumor microenvironment. Since BCRs/Igs in tumor tissues have high affinities for tumor-specific antigens, the patterns of their amino acid sequences and other sequence-independent features such as the number of somatic hypermutations (SHMs) may differ between the normal and tumor microenvironments. However, given the high diversity of BCRs/Igs and the rarity of recurrent sequences among individuals, it is far more difficult to capture such differences in BCR/Ig sequences than in TCR sequences. The aim of this study was to explore the possibility of discriminating BCRs/Igs in tumor and in normal tissues, by capturing these differences using supervised machine learning methods applied to RNA sequences of BCRs/Igs.\n\nRNA sequences of BCRs/Igs were obtained from matched normal and tumor specimens from 90 gastric cancer patients. BCR/Ig-features obtained in Rep-Seq were used to classify individual BCR/Ig sequences into normal or tumor classes. Different machine learning models using various features were constructed as well as gradient boosting machine (GBM) classifier combining these models. The results demonstrated that BCR/Ig sequences between normal and tumor microenvironments exhibit their differences. Next, by using a GBM trained to classify individual BCR/Ig sequences, we tried to classify sets of BCR/Ig sequences into normal or tumor classes. As a result, an area under the curve (AUC) value of 0.826 was achieved, suggesting that BCR/Ig repertoires have distinct sequence-level features in normal and tumor tissues.\n\nTo the best of our knowledge, this is the first study to show that BCR/Ig sequences derived from tumor and normal tissues have globally distinct patterns, and that these tissues can be effectively differentiated using BCR/Ig repertoires.

bioinformatics↗