Most downloaded biology preprints, all time
in category animal behavior and cognition
1,718 results found. For more information, click each entry to expand.
3,276 downloads bioRxiv animal behavior and cognition
Scientific datasets are growing rapidly in scale and complexity. Consequently, the task of understanding these data to answer scientific questions increasingly requires the use of compression algorithms that reduce dimensionality by combining correlated features and cluster similar observations to summarize large datasets. Here we introduce a method for both dimension reduction and clustering called VAE-SNE (variational autoencoder stochastic neighbor embedding). Our model combines elements from deep learning, probabilistic inference, and manifold learning to produce interpretable compressed representations while also readily scaling to tens-of-millions of observations. Unlike existing methods, VAE-SNE simultaneously compresses high-dimensional data and automatically learns a distribution of clusters within the data \---| without the need to manually select the number of clusters. This naturally creates a multi-scale representation, which makes it straightforward to generate coarse-grained descriptions for large subsets of related observations and select specific regions of interest for further analysis. VAE-SNE can also quickly and easily embed new samples, detect outliers, and can be optimized with small batches of data, which makes it possible to compress datasets that are otherwise too large to fit into memory. We evaluate VAE-SNE as a general purpose method for dimensionality reduction by applying it to multiple real-world datasets and by comparing its performance with existing methods for dimensionality reduction. We find that VAE-SNE produces high-quality compressed representations with results that are on par with existing nonlinear dimensionality reduction algorithms. As a practical example, we demonstrate how the cluster distribution learned by VAE-SNE can be used for unsupervised action recognition to detect and classify repeated motifs of stereotyped behavior in high-dimensional timeseries data. Finally, we also introduce variants of VAE-SNE for embedding data in polar (spherical) coordinates and for embedding image data from raw pixels. VAE-SNE is a robust, feature-rich, and scalable method with broad applicability to a range of datasets in the life sciences and beyond. ### Competing Interest Statement The authors have declared no competing interest.
3,174 downloads bioRxiv animal behavior and cognition
Rodent defense behavior assays have been widely used as preclinical models of anxiety to study possibly therapeutic anxiety-reducing interventions. However, some proposed anxiety-modulating factors - genes, drugs and stressors - have had discordant effects across different studies. To reconcile the effect sizes of purported anxiety factors, we conducted systematic review and meta-analyses of the literature on ten anxiety-linked interventions, as examined in the elevated plus maze, open field and light-dark box assays. Diazepam, 5-HT1A receptor gene knockout and overexpression, SERT gene knockout and overexpression, pain, restraint, social isolation, corticotropin-releasing hormone and Crhr1 were selected for review. Eight interventions had statistically significant effects on rodent anxiety, while Htr1a overexpression and Crh knockout did not. Evidence for publication bias was found in the diazepam, Htt knockout, and social isolation literatures. The Htr1a and Crhr1 results indicate a disconnect between preclinical science and clinical research. Furthermore, the meta-analytic data confirmed that genetic SERT anxiety effects were paradoxical in the context of the clinical use of SERT inhibitors to reduce anxiety.
3,086 downloads bioRxiv animal behavior and cognition
Aberrant social behavior is a core feature of many neuropsychiatric disorders, yet the study of complex social behavior in freely moving rodents is relatively infrequently incorporated into preclinical models. This likely contributes to limited translational impact. A major bottleneck for the adoption of socially complex, ethology-rich, preclinical procedures are the technical limitations for consistently annotating detailed behavioral repertoires of rodent social behavior. Manual annotation is subjective, prone to observer drift, and extremely time-intensive. Commercial approaches are expensive and inferior to manual annotation. Open-source alternatives often require significant investments in specialized hardware and significant computational and programming knowledge. By combining recent computational advances in convolutional neural networks and pose-estimation with further machine learning analysis, complex rodent social behavior is primed for inclusion under the umbrella of computational neuroethology. Here we present an open-source package with graphical interface and workflow (Simple Behavioral Analysis, SimBA) that uses pose-estimation to create supervised machine learning predictive classifiers of rodent social behavior, with millisecond resolution and accuracies that can out-perform human observers. SimBA does not require specialized video acquisition hardware nor extensive computational background. Standard descriptive statistical analysis, along with graphical region of interest annotation, are provided in addition to predictive classifier generation. To increase ease-of-use for behavioural neuroscientists, we designed SimBA with accessible menus for pre-processing videos, annotating behavioural training datasets, selecting advanced machine learning options, robust classifier validation functions and flexible visualizations tools. This allows for predictive classifier transparency, explainability and tunability prior to, and during, experimental use. We demonstrate that this approach is flexible and robust in both mice and rats by classifying social behaviors that are commonly central to the study of brain function and social motivation. Finally, we provide a library of pose-estimation weights and behavioral predictive classifiers for resident-intruder behaviors in mice and rats. All code and data, together with detailed tutorials and documentation, are available on the SimBA GitHub repository. ### Competing Interest Statement The authors have declared no competing interest.
2,824 downloads bioRxiv animal behavior and cognition
Progress in understanding how individual animals learn will require high-throughput standardized methods for behavioral training but also advances in the analysis of the resulting behavioral data. In the course of training with multiple trials, an animal may change its behavior abruptly, and capturing such events calls for a trial-by-trial analysis of the animal's strategy. To address this challenge, we developed an integrated platform for automated animal training and analysis of behavioral data. A low-cost and space-efficient apparatus serves to train entire cohorts of mice on a decision-making task under identical conditions. A generalized linear model (GLM) analyzes each animal's performance at single-trial resolution. This model infers the momentary decision-making strategy and can predict the animal's choice on each trial with an accuracy of ~80%. We also assess the animal's detailed trajectories and body poses within the apparatus. Unsupervised analysis of these features revealed unusual trajectories that represent hesitation in the response. This integrated hardware/software platform promises to accelerate the understanding of animal learning.
2,709 downloads bioRxiv animal behavior and cognition
From foraging for food to learning complex games, many aspects of human behaviour can be framed as a search problem with a vast space of possible actions. Under finite search horizons, optimal solutions are generally unobtainable. Yet how do humans navigate vast problem spaces, which require intelligent exploration of unobserved actions? Using a variety of bandit tasks with up to 121 arms, we study how humans search for rewards under limited search horizons, where the spatial correlation of rewards (in both generated and natural environments) provides traction for generalization. Across a variety of different probabilistic and heuristic models, we find evidence that Gaussian Process function learning--combined with an optimistic Upper Confidence Bound sampling strategy--provides a robust account of how people use generalization to guide search. Our modelling results and parameter estimates are recoverable, and can be used to simulate human-like performance, providing novel insights about human behaviour in complex environments.
2,665 downloads bioRxiv animal behavior and cognition
Habitual reliance on tool use is a marked behavioral difference between wild robust (genus Sapajus) and gracile (genus Cebus) capuchin monkeys. Despite being well studied and having a rich repertoire of social and extractive foraging traditions, Cebus sp have rarely been observed engaging in tool use and have never been reported to use stone tools. In contrast, habitual tool use and stone-tool use by Sapajus is widespread. We discuss factors which might explain these differences in patterns of tool use between Cebus and Sapajus. We then report the first case of habitual stone-tool use in a gracile capuchin: a population of white-faced capuchins (Cebus capucinus imitator) in Coiba National Park, Panama who habitually rely on hammerstone and anvil tool use to access structurally protected food items in coastal areas including Terminalia catappa} seeds, hermit crabs, marine snails, terrestrial crabs, and other items. This behavior has persisted on one island in Coiba National Park since at least 2004. From one year of camera trapping, we found that stone tool use is strongly male-biased. Of the 205 unique camera-trap-days where tool use was recorded, adult females were never observed to use stone-tools, although they were frequently recorded at the sites and engaged in scrounging behavior. Stone-tool use occurs year-round in this population, and over half of all identifiable individuals were observed participating. At the most active tool use site, 83.2% of days where capuchins were sighted corresponded with tool use. Capuchins inhabiting the Coiba archipelago are highly terrestrial, under decreased predation pressure and potentially experience resource limitation compared to mainland populations-- three conditions considered important for the evolution of stone tool use. White-faced capuchin tool use in Coiba National Park thus offers unique opportunities to explore the ecological drivers and evolutionary underpinnings of stone tool use in a comparative within- and between-species context.
2,588 downloads bioRxiv animal behavior and cognition
Characteristics of male courtship behavior in Drosophila melanogaster have been well-described, but the genetic basis of male-female copulation is largely unknown. Here we show that the white (w) gene, a classical gene for eye color, is associated with copulation success. 82.5% of wild-type Canton-S flies copulated within 60 minutes in circular arenas, whereas few white-eyed mutants mated successfully. The w+ allele exchanged to the X chromosome or duplicated to the Y chromosome in the white-eyed genetic background rescued the defect of copulation success. The w+-associated copulation success was independent of eye color phenotype. Addition of the mini-white (mw+) gene to the white-eyed mutant rescued the defect of copulation success in a manner that was mw+ copy number-dependent. Lastly, male-female sexual experience mimicked the effects of w+/mw+ in improving successful copulation. These data suggest that the w+ gene controls copulation success in Drosophila melanogaster.
2,525 downloads bioRxiv animal behavior and cognition
Dogs can interpret emotional human faces (especially the ones expressing happiness), yet the cerebral correlates of this process are unknown. Using functional magnetic resonance imaging (fMRI) we studied eight awake and unrestrained dogs. In Experiment 1 dogs observed happy and neutral human faces, and found increased brain activity when viewing happy human faces in temporal cortex and caudate. In Experiment 2 the dogs were presented with human faces expressing happiness, anger, fear, or sadness. Using the resulting cluster from Experiment 1 we trained a linear support vector machine classifier to discriminate between pairs of emotions and found that it could only discriminate between happiness and the other emotions. Finally, evaluation of the whole-brain fMRI time courses through a similar classifier allowed us to predict the emotion being observed by the dogs. Our results show that human emotions are specifically represented in dogs' brains, highlighting their importance for inter-species communication.
2,520 downloads bioRxiv animal behavior and cognition
Nervous systems have evolved to combine environmental information with internal state to select and generate adaptive behavioral sequences. To better understand these computations and their implementation in neural circuits, natural behavior must be carefully measured and quantified. Here, we collect high spatial resolution video of single zebrafish larvae swimming in a naturalistic environment and develop models of their action selection across exploration and hunting. Zebrafish larvae swim in punctuated bouts separated by longer periods of rest called interbout intervals. We take advantage of this structure by categorizing bouts into discrete types and representing their behavior as labeled sequences of bout-types emitted over time. We then construct probabilistic models - specifically, marked renewal processes - to evaluate how bout-types and interbout intervals are selected by the fish as a function of its internal hunger state, behavioral history, and the locations and properties of nearby prey. Finally, we evaluate the models by their predictive likelihood and their ability to generate realistic trajectories of virtual fish swimming through simulated environments. Our simulations capture multiple timescales of structure in larval zebrafish behavior and expose many ways in which hunger state influences their action selection to promote food seeking during hunger and safety during satiety.
2,505 downloads bioRxiv animal behavior and cognition
In order to discover the most rewarding actions, agents must collect information about their environment, potentially foregoing reward. The optimal solution to this "explore-exploit" dilemma is often computationally challenging, but principled algorithmic approximations exist. These approximations utilize uncertainty about action values in different ways. Some random exploration algorithms scale the level of choice stochasticity with the level of uncertainty. Other directed exploration algorithms add a "bonus" to action values with high uncertainty. Random exploration algorithms are sensitive to total uncertainty across actions, whereas directed exploration algorithms are sensitive to relative uncertainty. This paper reports a multi-armed bandit experiment in which total and relative uncertainty were orthogonally manipulated. We found that humans employ both exploration strategies, and that these strategies are independently controlled by different uncertainty computations.
2,495 downloads bioRxiv animal behavior and cognition
We present an open-source software platform that transforms the emotions expressed by speech signals using audio effects like pitch shifting, inflection, vibrato, and filtering. The emotional transformations can be applied to any audio file, but can also run in real-time (with less than 20-millisecond latency), using live input from a microphone. We anticipate that this tool will be useful for the study of emotions in psychology and neuroscience, because it enables a high level of control over the acoustical and emotional content of experimental stimuli in a variety of laboratory situations, including real-time social situations. We present here results of a series of validation experiments showing that transformed emotions are recognized at above-chance levels in the French, English, Swedish and Japanese languages, with a naturalness comparable to natural speech. Then, we provide a list of twenty-five experimental ideas applying this new tool to important topics in the behavioral sciences.
2,470 downloads bioRxiv animal behavior and cognition
To restore vision for the blind several prosthetic approaches have been explored that convey raw images to the brain. So far these schemes all suffer from a lack of bandwidth and the extensive training required to interpret unusual stimuli. Here we present an alternate approach that restores vision at the cognitive level, bypassing the need to convey sensory data. A wearable computer captures video and other data, extracts the important scene knowledge, and conveys that through auditory augmented reality. This system supports many aspects of visual cognition: from obstacle avoidance to formation and recall of spatial memories, to long-range navigation. Neither training nor modification of the physical environment are required: Blind subjects can navigate an unfamiliar multi-story building on their first attempt. The combination of unprecedented computing power in wearable devices with augmented reality technology promises a new era of non-invasive prostheses that are limited only by software.
2,467 downloads bioRxiv animal behavior and cognition
Information on genetic relationships among individuals is essential to many studies of the behavior and ecology of wild organisms. Parentage and relatedness assays based on large numbers of SNP loci hold substantial advantages over the microsatellite markers traditionally used for these purposes. We present a double-digest restriction site-associated DNA sequencing (ddRAD-seq) analysis pipeline that, as such, simultaneously achieves the SNP discovery and genotyping steps and which is optimized to return a statistically powerful set of SNP markers (typically 150-600 after stringent filtering) from large numbers of individuals (up to 240 per run). We explore the tradeoffs inherent in this approach through a set of experiments in a species with a complex social system, the variegated fairy-wren (Malurus lamberti), and further validate it in a phylogenetically broad set of other bird species. Through direct comparisons with a parallel dataset from a robust panel of highly variable microsatellite markers, we show that this ddRAD-seq approach results in substantially improved power to discriminate among potential relatives and considerably more precise estimates of relatedness coefficients. The pipeline is designed to be universally applicable to all bird species (and with minor modifications to many other taxa), to be cost- and time-efficient, and to be replicable across independent runs such that genotype data from different study periods can be combined and analyzed as field samples are accumulated.
2,467 downloads bioRxiv animal behavior and cognition
Mainstream cognitive science and neuroscience both rely heavily on the notion of representation in order to explain the full range of our behavioral repertoire. The relevant feature of representation is its ability to designate (stand in for) spatially or temporally distant properties, When we organize our behavior with respect to mental or neural representations, we are (in principle) organizing our behavior with respect to the property it designates. While representational theories are a potentially a powerful foundation for a good cognitive theory, problems such as grounding and system-detectable error remain unsolved. For these and other reasons, ecological explanations reject the need for representations and do not treat the nervous system as doing any mediating work. However, this has left us without a straight-forward vocabulary to engage with so-called 'representation-hungry' problems or the role of the nervous system in cognition. In an effort to develop such a vocabulary, here we show that James J Gibson's ecological information functions to designate the ecologically-scaled dynamical world to an organism. We then show that this designation analysis of information leads to an ecological conceptualization of the neural activity caused by information, which in turn we argue can together support intentional behavior with respect to spatially and temporally distal properties. Problems such as grounding and error detection are solved via law-based specification. This analysis extends the ecological framework into the realm of 'representation-hungry' problems, making it as powerful a potential basis for theories of behavior as traditional cognitive approaches. The resulting analysis does, according to some definitions, allow information and the neural activity to be conceptualized as representations; however, the key work is done by information and the analysis remains true to Gibson's ecological ontology.
2,380 downloads bioRxiv animal behavior and cognition
In most mammalian species, females regularly interact with kin, and it may thus be difficult to understand the evolution of some aggressive and harmful competitive behaviour among females, such as infanticide. Here, we investigate the evolutionary determinants of infanticide by females by combining a quantitative analysis of the taxonomic distribution of infanticide with a qualitative synthesis of the circumstances of infanticidal attacks in published reports. Our results show that female infanticide is widespread across mammals and varies in relation to social organization and life-history, being more frequent where females breed in groups and have intense bouts of high reproductive output. Specifically, female infanticide occurs where the proximity of conspecific offspring directly threatens the killer's reproductive success by limiting access to critical resources for her dependent progeny, including food, shelters, care or a social position. In contrast, infanticide is not immediately modulated by the degree of kinship among females, and females occasionally sacrifice closely related juveniles. Our findings suggest that the potential direct fitness rewards of gaining access to reproductive resources have a stronger influence on the expression of female aggression than the indirect fitness costs of competing against kin.
2,113 downloads bioRxiv animal behavior and cognition
A goal of computational psychiatry is to ground symptoms in more fundamental computational mechanisms. Theory suggests that rumination and other symptoms in mood disorders reflect dysregulated mental simulation, a process that normally serves to evaluate candidate actions. If so, these covert symptoms should have observable consequences: excessively deliberative choices, specifically about options related to the content of rumination. In two large general population samples, we examined how symptoms of social anxiety disorder (SAD) predict choices in a socially framed reinforcement learning task, the Patent Race game. Using a computational learning model to assess learning strategy, we found that self-reported social anxiety was indeed associated with an increase in deliberative evaluation. The effect was specific to learning from a particular ("upward counterfactual") subset of feedback, broadly matching the biased content of rumination in SAD. It was also robust to controlling for other psychiatric symptoms. These results ground the symptoms of SAD, such as overthinking and paralysis in social interactions, in well characterized neuro-computational mechanisms and offer a rare example of enhanced function in disease.
2,093 downloads bioRxiv animal behavior and cognition
The exploration-exploitation dilemma is one of a few fundamental problems in the decision and life sciences. It has also been proven to be a mathematically intractable problem, when it is framed in terms of reward. To overcome this we have challenged the basic formulation of the problem itself, as having a single objective, namely reward. In its place we have defined independent objectives for exploration and exploitation. Through theory and numerical experiments we prove that a competition between exploration-as-curiosity and exploitation-for-reward has a tractable solution, based on the classic win-stay, lose-switch strategy from game theory. This strategy is possible because we treat information and reward as if they have equal value, and succeeds because the definition of curiosity we introduce is efficient. Besides offering a mathematical answer, this view of the problem seems more robust than the traditional approach because it succeeds in the difficult conditions where rewards are deceptive, or non-stationary.
2,076 downloads bioRxiv animal behavior and cognition
Automated movement tracking is essential for high-throughput quantitative analyses of the behaviour and kinematics of organisms. Automated tracking also improves replicability by avoiding observer biases and allowing reproducible workflows. However, few automated tracking programs exist that are open access, open source, and capable of tracking unmarked organisms in noisy environments. Tracktor is an image-based tracking freeware designed to perform single-object tracking in noisy environments, or multi-object tracking in uniform environments while maintaining individual identities. Tracktor is code-based but requires no coding skills other than the user being able to specify tracking parameters in a designated location, much like in a graphical user interface (GUI). The installation and use of the software is fully detailed in a user manual. Through four examples of common tracking problems, we show that Tracktor is able to track a variety of animals in diverse conditions. The main strengths of Tracktor lie in its ability to track single individuals under noisy conditions (e.g. when the object shape is distorted), its robustness to perturbations (e.g. changes in lighting conditions during the experiment), and its capacity to track multiple individuals while maintaining their identities. Additionally, summary statistics and plots allow measuring and visualizing common metrics used in the analysis of animal movement (e.g. cumulative distance, speed, acceleration, activity, time spent in specific areas, distance to neighbour, etc.). Tracktor is a versatile, reliable, easy-to-use automated tracking software that is compatible with all operating systems and provides many features not available in other existing freeware. Access Tracktor and the complete user manual here: https://github.com/vivekhsridhar/tracktor
2,047 downloads bioRxiv animal behavior and cognition
Domestic dogs are remarkably sensitive and responsive while interacting with humans. Pet dogs are known to have social skills and abilities to display situation-specific responses, but there is lack of information regarding free-ranging dogs which constitute majority of the world's dog population. Free-ranging dogs found in most of the developing countries interact constantly with familiar and unfamiliar humans receiving both positive and negative behavior. Thus, understanding human intentions and subsequent behavioral adjustments are crucial for dogs that share habitats with humans. Here we subjected free-ranging dogs to different human social communicative cues, followed by a food provisioning phase and tested their responsiveness. Dogs exhibited higher proximity seeking behavior as a reaction to friendly gesture whereas, they were prompted to maintain distance depending on the impact of the threatening cues. Interestingly, only the high-impact threatening showed to have a persistent effect which also remained during the subsequent food provisioning phase. An elevated approach in the food provisioning phase elicited the dependency of free-ranging dogs on humans for sustenance. Our findings suggest that free-ranging dogs demonstrate behavioral plasticity on interacting with humans; which provides significant insights into the establishment of the dog-human relationship on streets.
2,041 downloads bioRxiv animal behavior and cognition
To study brain function, preclinical research relies heavily on animal monitoring and the subsequent analyses of behavior. Commercial platforms have enabled semi high-throughput behavioral analyses by providing accurate tracking of animals, yet they often struggle with the analysis of ethologically relevant behaviors and lack the flexibility to adapt to variable testing environments. In the last couple of years, substantial advances in deep learning and machine vision have given researchers the ability to take behavioral analysis entirely into their own hands. Here, we directly compare the performance of commercially available platforms (Ethovision XT14, Noldus; TSE Multi Conditioning System, TSE Systems) to cross-verified human annotation. To this end, we provide a set of videos - carefully annotated by several human raters - of three widely used behavioral tests (open field, elevated plus maze, forced swim test). Using these data, we show that by combining deep learning-based motion tracking (DeepLabCut) with simple post-analysis, we can track animals in a range of classic behavioral tests at similar or even greater accuracy than commercial behavioral solutions. In addition, we integrate the tracking data from DeepLabCut with post analysis supervised machine learning approaches. This combination allows us to score ethologically relevant behaviors with similar accuracy to humans, the current gold standard, thus outperforming commercial solutions. Moreover, the resulting machine learning approach eliminates variation both within and between human annotators. In summary, our approach helps to improve the quality and accuracy of behavioral data, outperforming commercial systems at a fraction of the cost.
- 27 Nov 2020: The website and API now include results pulled from medRxiv as well as bioRxiv.
- 18 Dec 2019: We're pleased to announce PanLingua, a new tool that enables you to search for machine-translated bioRxiv preprints using more than 100 different languages.
- 21 May 2019: PLOS Biology has published a community page about Rxivist.org and its design.
- 10 May 2019: The paper analyzing the Rxivist dataset has been published at eLife.
- 1 Mar 2019: We now have summary statistics about bioRxiv downloads and submissions.
- 8 Feb 2019: Data from Altmetric is now available on the Rxivist details page for every preprint. Look for the "donut" under the download metrics.
- 30 Jan 2019: preLights has featured the Rxivist preprint and written about our findings.
- 22 Jan 2019: Nature just published an article about Rxivist and our data.
- 13 Jan 2019: The Rxivist preprint is live!