SoftwareX

Automated detection of complex animal behavior remains a challenge in neuroscience. Developments in computer vision have greatly advanced automated behavior detection and allow high-throughput preclinical and mechanistic studies. An integrated hardware and software solution is necessary to facilitate the adoption of these advances in the field of behavioral neurogenetics, particularly for non-computational laboratories. We have published a series of papers using an open field arena to annotate complex behaviors such as grooming, posture, and gait as well as higher-level constructs such as biological age and pain. Here, we present our, integrated rodent phenotyping platform, JAX Animal Behavior System (JABS), to the community for data acquisition, machine learning-based behavior annotation and classification, classifier sharing, and genetic analysis. The JABS Data Acquisition Module (JABS-DA) enables uniform data collection with its combination of 3D hardware designs and software for real-time monitoring and video data collection. JABS-Active Learning Module (JABS-AL) allows behavior annotation, classifier training, and validation. We introduce a novel graph-based framework (ethograph) that enables efficient boutwise comparison of JABS-AL classifiers. JABS-Analysis and Integration Module (JABS-AI), a web application, facilitates users to deploy and share any classifier that has been trained on JABS, reducing the effort required for behavior annotation. It supports the inference and sharing of the trained JABS classifiers and downstream genetic analyses (heritability and genetic correlation) on three curated datasets spanning 168 mouse strains that we are publicly releasing alongside this study. This enables the use of genetics as a guide to proper behavior classifier selection. This open-source tool is an ecosystem that allows the neuroscience and genetics community for shared advanced behavior analysis and reduces the barrier to entry into this new field.

Open source analytical software for the analysis of electrophysiological cardiomyocyte data offers a variety of new functionalities to complement closed-source, proprietary tools. Here, we present the Cardio PyMEA application, a free, modifiable, and open source program for the analysis of microelectrode array (MEA) data obtained from cardiomyocyte cultures. Cardio PyMEA was written entirely in Python 3 to provide an accessible, integrated workflow that possesses a user-friendly graphical user interface (GUI) written in PyQt5 to allow for performant, cross-platform utilization. This application makes use of object-oriented programming (OOP) principles to facilitate the relatively straightforward incorporation of custom functionalities, e.g. power law analysis, that suit the needs of the user. Cardio PyMEA is available as an open source application under the terms of the GNU General Public License (GPL). The source code for Cardio PyMEA can be downloaded from Github at the following repository: https://github.com/csdunhamUC/cardio_pymea.

Fluctuations and propagation of cytosolic calcium levels at both the cellular and tissue levels show complex patterns, referred to as calcium signatures, that regulate growth, organ development, damage responses, and survival. The quantitative analysis of calcium signatures at the cellular level is essential for identifying unique patterns that coordinate biological processes. However, a versatile framework applicable to multiple tissue types, allowing researchers to compare, measure, and validate diverse responses and recognize conserved patterns across model organisms, is missing. Here, we present a post-processing tool, CalciumInsights, which leverages the R packages Shiny and Golem. This tool has a graphical user interface and does not require software programming experience to perform calcium signal analysis. The open-source software has a modular framework with standardized functionalities that can be tailored for various research approaches. CalciumInsights provides descriptive statistical analysis through various metrics extracted from dynamic calcium transients and oscillations, such as peak amplitude, area under the curve, frequency, among others. The tool was evaluated with fluorescence imaging data from three model organisms: Danio rerio, Arabidopsis thaliana, and Drosophila melanogaster, demonstrating its ability to analyze diverse biological responses and models. Finally, the open-source nature of CalciumInsights enables community-driven improvements and developments for enabling new applications. Author SummaryThis manuscript introduces CalciumInsights, an open-source tool for calcium signature analysis. Designed to be a versatile tool that works with various tissue types and biological systems, CalciumInsights has an easy-to-use graphical user interface. Our program simplifies metrics extraction while maintaining the quality of the analysis by integrating several algorithms. CalciumInsights stands out for its user-friendliness, ease of use, and robust data exploration features, such as tunable filters for improved accuracy. These features promote inclusivity and lower barriers to scientific research by making calcium signature analysis accessible to users of all programming skill levels.

Non-negative matrix factorization (NMF) produces a factorization that constrains the elements of both the factor matrices to be non-negative. It has been a popular feature extraction method in many applications including neuroimaging. One limitation of the existing softwares of NMF is that they were written in and dependent on the proprietary software of MATLAB. To address this limitation, we introduced an open-source C++ package for performing NMF. To make NMF more accessible to the scientific research community, we describe a NMF algorithm implemented using the Insight Toolkit ITK and Armadillo, a MATLAB style C++ based math library. Armadillo facilitates the computations in linear algebra by calling functions without any need to implement functions in C++. In addition, This framework supports the read and write interface to images specific to neuroscience. Finally, The package supports NMF with multiplicative update and sophisticated initialization methods. We showed that the package has accuracy matching MATLAB and its speed close to that of MATLAB. We used simple simulated images to test its functionality. Then, we demonstrated how the package can be used to analyze neuroimaging data. Specifically, we used the package to find a data-driven set of structural patterns(factor matrices) that are similar across individuals. We validated this factorization method by associating their weighted loading matrices with body mass indices (BMI) of individuals from the human connectome project.

AO_SCPLOWBSTRACTC_SCPLOWHead-mounted tools for eye/head tracking are increasingly used for assessment of visual behavior in navigation, sports, sociology, and neuroeconomics. Here we introduce an open-source python software (TP3Py) for collection and analysis of portable eye/head tracking signals using Tobii Pro Glasses 3. TP3Pys modular pipeline provides a platform for incorporating user-oriented functionalities and comprehensive data acquisition to accelerate the development in behavioral and tracking research. Tobii Pro Glasses 3 is equipped with embedded cameras viewing the visual scene and the eyes, inertial measurement unit (IMU) sensors, and video-based eye tracker implemented in the accompanying unit. The program establishes a wireless connection to the glasses and, within separate threads, continuously leverages the received data in numerical or string formats accessible for saving, processing, and graphical purposes. Built-in modules for presenting eye, scene, and IMU data to the experimenter have been adapted as well as communicating modules for sending the raw signals to stimulus/task controllers in live fashion. Closed-loop experimental designs are limited due to the 140ms time delay of the system, but this limitation is compensated by the portability of the eye/head tracking. An offline data viewer has been also incorporated to allow more time-consuming computations. Lastly, we demonstrate example recordings involving vestibulo-ocular reflexes, saccadic eye movements, optokinetic responses, or vergence eye movements to highlight the programs measurement capabilities to address various experimental goals. TP3Py has been tested on Windows with Intel processors, and Ubuntu operating systems with Intel or ARM (Raspberry Pie) architectures.

AbstractGait assessments are a key part of determining the wellbeing of livestock. Techniques for gait assessment have traditionally involved human-eye inspections or reflective markers, but markerless computer vision methods have been developed in recent years. Despite many computer vision tools providing high-quality pose estimations in an efficient manner, they lack post-processing functionality. A review of model performance and calculation of gait parameters is a necessary step to fully harness the capability of this new technology. Thus, this study developed DeepLabCut-Display, an open-source desktop software application. DeepLabCut-Display allows a user to upload the video and coordinate data associated with the output of DeepLabCut, a prominent pose-estimation software tool. A user can review the video and coordinate data in parallel, filter points by a likelihood threshold, and automatically calculate gait parameters. Specific video frames, filtered data, and gait parameters can be exported from the application for further usage. The source code is publicly hosted on GitHub alongside installation and usage instructions. DeepLabCut-Display, the product of interdisciplinary and collaborative design between software developers and animal scientists, will alleviate a critical bottleneck in processing of data for locomotor analysis in livestock. Summary StatementO_LIDeepLabCut-Display is a utility to dynamically visualize raw marker coordinates, and to automatically produce gait parameters for locomotion analysis of horses and other livestock. C_LI Lay SummaryArtificial intelligence systems that can predict and track the positions of objects are now being applied in many fields, including animal science. Veterinarians and animal scientists use these systems to create pose estimations, a digital label of anatomical landmarks overlaid on a video of an animal in motion. They are used to quantify the subjects motion and detect anomalies that may be indicative of disease or injury. Pose estimation systems are efficient and accurate, but they lack features like data visualization and post-processing analysis that are necessary to make determinations about the animals motion. This study developed DeepLabCut-Display, a software application that can visualize the data from a pose estimation system and provides a set of tools for further analysis. After a user is done with analysis, they can save the results to their computer. The application was made by a collaboration between software developers and animal scientists, highlighting how interdisciplinary teams are effective at producing useful software.

Conformational sampling of complex biomolecules is an emerging frontier in drug discovery. Indeed, advances in lab-based structural biology and related computational approaches like AlphaFold have made great strides in obtaining static protein structures. However, biology is in constant motion and many important biological processes rely on conformationally-driven events. Unrestrained molecular dynamics (MD) simulations require that the simulated time be comparable to the real time of the biological processes of interest, rendering pure MD impractical for many drug design projects, where conformationally-driven biological events can take microseconds to milliseconds or longer. An alternative approach is to accelerate the sampling of specific motions by applying restraints, guided by insights about the underlying biological process of interest. A plethora of restraints exist to limit the size of conformational search space, although each has drawbacks when simulating complex biological motions. In this work, we introduce a new kind of restraint for molecular dynamics simulations (MD) that is particularly well suited for complex conformationallydriven biological events, such as protein-ligand binding, allosteric modulations, conformational signalling, and membrane permeability. The new restraint, which relies on a barrier function (the scaled reciprocal function) is particularly beneficial to MD, where hard-wall restraints are needed with zero tolerance to restraint violation. We have implemented this restraint within a hybrid sampling framework that combines metadynamics and extended-Lagrangian adaptive biasing force (meta-eABF). We use two particular examples to demonstrate the value of this approach: (1) quantification of the approach of E3-loaded ubiquitin to a protein of interest as part of the Cullin ring ligase and (2) membrane permeability of heterobi-functional degrader molecules with a large degree of conformational flexibility. Future work will involve extension to additional systems and benchmarking of this approach compared with other methods.

Single Particle Tracking (SPT) is one of the most widespread techniques to evaluate particle mobility in a variety of situations, such as in cellular and model membrane dynamics. The proposed TRAIT2D Python library is developed to provide object tracking, trajectory analysis and produce simulated datasets with graphical user interface. The tool allows advanced users to customise the analysis to their requirements. Availability and implementation: the software has been coded in Python, and can be accessed from: https://github.com/Eggeling-Lab-Microscope-Software/TRAIT2D, or the pypi and condaforge repositories. A comprehensive user guide is provided at https://eggeling-lab-microscope-software.github.io/TRAIT2D/.

Calcium signaling data analysis has become increasing complex as the size of acquired datasets increases. In this paper we present a Ca2+ signaling data analysis method that employs custom written software scripts deployed in a collection of Jupyter-Lab "notebooks" which were designed to cope with this complexity. The notebook contents are organized to optimize data analysis workflow and efficiency. The method is demonstrated through application to several different Ca2+ signaling experiment types.

Graph data is an omnipresent way to represent information in machine learning. Especially, in neuroscience research, data from Diffusion-Tensor Imaging (DTI) and functional Magnetic Resonance Imaging (fMRI) is commonly represented as graphs. Exploiting the graph structure of these modalities using graph-specific machine learning applications is currently hampered by the lack of easy-to-use software. PHOTONAI Graph aims to close the gap between domain experts of machine learning, graph experts and neuroscientists. Leveraging the rapid machine learning model development features of the Python machine learning API PHOTONAI, PHOTONAI Graph enables the design, optimization, and evaluation of reliable graph machine learning models for practitioners. As such, it provides easy access to custom graph machine learning pipelines including, hyperparameter optimization and algorithm evaluation ensuring reproducibility and valid performance estimates. Integrating established algorithms such as graph neural networks, graph embeddings and graph kernels, it allows researchers without significant coding experience to build and optimize complex graph machine learning models within a few lines of code. We showcase the versatility of this toolbox by building pipelines for both resting-state fMRI and DTI data in the hope that it will increase the adoption of graph-specific machine learning algorithms in neuroscience research.

Chaste is an open-source C++ library providing a general-purpose framework for cell-based simulations of biological tissues. It has been applied to a wider range of biological processes, including morphogenesis, carcinogenesis, and wound healing. Such simulations often involve numerous mechanical interactions between neighbouring cells, making them computationally demanding. Graphical Processing Units (GPUs), with their highly parallel architectures, offer a powerful means to accelerate these computations, enabling larger, more detailed simulations and improving research productivity. FLAME GPU 2 is a GPU-accelerated simulator for domain-independent complex systems that maps formal agent descriptions written in scripting language to optimized CUDA code. In this work, FLAME GPU 2 is integrated with Chaste to accelerate force calculations in a class of cell-based simulations, demonstrating the feasibility of GPU acceleration within existing CPU-based frameworks. The GPU-accelerated implementation is validated against the original CPU version, achieving up to 93.6x speedup in force calculations and 3.72x speedup for full simulations across various cell population sizes. Moreover, the approach enables smaller mechanics time steps, without incurring significant data transfer overhead, thereby improving the accuracy of mechanical modelling. This enhancement increases the fidelity of cell position calculations in non-equilibrium simulations and improves dynamic accuracy as cells approach equilibrium.

The historically fragmented biomedical data ecosystem has moved towards harmonization under the findable, accessible, interoperable, and reusable (FAIR) data principles, creating more opportunities for cloud-based research. This shift is especially opportune for scientists across diverse domains interested in implementing creative, nonstandard computational analytic pipelines on large and varied datasets. However, executing custom cloud analyses may present difficulties, particularly for investigators lacking advanced computational expertise. Here, we present an accessible, streamlined approach for the cloud compute platform CAVATICA that offers a solution. We outline how we developed a custom workflow in the cloud, for analyzing whole genome sequences of case-parent trios to detect sex-specific genetic effects on orofacial cleft risk, which required several programming languages and custom software packages. The approach involves just three components: Docker to containerize software environments, tool creation for each analysis step, and a visual workflow editor to weave the tools into a Common Workflow Language (CWL) pipeline. Our approach should be accessible to any investigator with basic computational skills, is readily extended to implement any scalable high-throughput biomedical data analysis in the cloud, and is applicable to other commonly used compute platforms such as BioData Catalyst. We believe our approach empowers versatile data reuse and promotes accelerated biomedical discovery in a time of substantial FAIR data.

Considering the intricate nature of biological processes within organisms, it is undeniable that relying solely on in vitro-generated primary-cell-like cultures or organ-like products in preclinical and basic research is insufficient to replace animal-based studies fully. This limitation is particularly significant when considering the regulations enforced by legislative assemblies worldwide. The necessity of animal-based studies to approve chemicals and medications. In contradiction, European countries aim to banish animal-based studies. Therefore, we must understand the impact of the data refinement and experiment replacement strategy we will introduce here. This project aimed to revolutionize data acquisition in animal-based studies by transforming manual observation into a reliable digital process. Reliable digital data will be generated by having the potential to reduce human bias by simply reducing human interaction. Additionally, reducing human interaction will reduce the severity levels due to stress reduction, fulfilling the 3R principles. Therefore, the first goal was to develop and implement a scalable, stable, running, and remotely accessible camera-based monitor system (the iMouse solution). At the same time, the target was to develop a retrofit solution (DigiFrame) for existing home-cage systems, not interfering with the regular workflow in animal facilities. As a result, we developed a digital monitoring system, named iMouseTV platform based on existing open-source software, allowing users to observe, record, share, and review animal-based studies within the home cage anytime from anywhere, reducing the stress level for the animals. Our systems first Proof of concept ran for over two years at the LIV in Hamburg. We also investigated an effective way to reduce data generation by setting up specific zones for detecting the motion of choice (e.g., drinking, food intake). The data sets can be stored, shared, and reviewed by users and refined by algorithms aiming to recognize the dedicated motions of the animals automatically. The implementation of the ML algorithms allows the iMouse solution to recognize whether an individual mouse was drinking and for how long and store results in the annotated video file and graph format. However, the identification and continuous tracking of the species is still in progress. In conclusion, we established a scalable human-independent monitoring and recording system, which can be implemented into the existing structures of institutions and companies without changing handling processes, to monitor animals and observe them by getting reliable digital data. Moreover, it is fundamental for automatic recognition within animal-based studies based on Artificial Intelligence.

As high-performance computing provides the ability to generate and analyze ever larger simulation trajectories, the challenges in learning, applying, and sharing the best analytical practices become more salient. Extracting reproducible scientific insights from simulation requires a thorough understanding of many computing topics unrelated to the molecular systems being modeled and simulated. While the rapid development of the technologies used for analysis makes previously impossible studies into routine work, the growing repertoire of software combined with the specificity of the ecosystems that they rely on can easily break the programs used in older studies. In this work, we present ST-Analyzer, a simulation trajectory analysis suite with command-line (CLI) and graphical (GUI) user interfaces. ST-Analyzer is distributed freely as an open-source conda-forge package with support for macOS, Linux, and Windows (via WSL2). Besides facilitating several common analysis tasks, the GUI shows users the exact commands necessary to repeat the same tasks on the command-line. We demonstrate ST-Analyzers capabilities by reproducing several results from previously published simulation studies on the lipid parameters of heterogeneous biomembranes and the behavior of a SARS-CoV-2 spike protein-antibody complex. We expect ST-Analyzer to be useful to experts for quickly setting up common analysis tasks and to nonexperts as a guided introduction to simulation analysis using both GUI and CLI. ST-Analyzer is freely available at https://github.com/nk53/stanalyzer.

The problem of finding the configuration of points given partial information on pairwise inter-point distances, the Euclidean distance geometry problem, appears in multiple applications. In this paper, we propose an approach that integrates structural similarity and a nonconvex distance geometry algorithm for the protein structure determination problem. When initialized with a homologous structure, reconstruction of ubiquitin structure with our non convex algorithm resulted in an RMSE of less than 2 [A] with 1.5% available inter proton distance and up to 20% relative error in the input distances. To test the robustness of this algorithm with regard to initialization, we also initialized with a nonhomologous structure on a larger protein with pdb coordinate 1W2E. Even though the initialization structure 1JYB is far different from 1W2E with an RMSE of 25 [A], reconstruction generated structures with RMSE of close to 2 [A], using 1.7% available proton distances and up to 10% relative error in input distances. These results suggest EDG-based approach may be applied to fast NMR structure determination in the future.

Fluorescence lifetime imaging microscopy (FLIM) translates the duration of excited states of fluorophores into lifetime information as additional source of contrast in images of biological samples. This offers the possibility to separate fluorophores particularly beneficial in case of similar excitation spectra. Here, we demonstrate the distinction of fluorescent molecules based on FLIM phasor analysis, called lifetime unmixing, in live-cell imaging using open-source software for analysis. We showcase two applications using Caenorhabditis elegans as a model system. First, we unmixed the highly spectrally overlapping fluorophores mCherry and mKate2 to distinctively track tagged proteins in six-dimensional datasets to investigate cell division in the developing early embryo. Second, we unmixed fluorescence of tagged proteins of interest from masking natural autofluorescence in adult hermaphrodites. For FLIM data handling and workflow implementation, we developed the open-source plugin napari-FLIM-phasor-plotter to implement conversion, visualization, analysis and reuse of FLIM data of different formats. Our work thus advances technical applications and bioimage data management and analysis in FLIM microscopy for life science research.

Density-based clustering procedures are widely used in a variety of data science applications. Their advantage lies in the capability to find arbitrarily shaped and sized clusters and robustness against outliers. In particular, they proved effective in the analysis of Molecular Dynamics simulations, where they serve to identify relevant, low energetic molecular conformations. As such, they can provide a convenient basis for the construction of kinetic (coreset) Markov-state models. Here we present the opensource Python project CommonNNClustering, which provides an easy-to-use and efficient re-implementation of the commonnearest-neighbour (CommonNN) method. The package provides functionalities for hierarchical clustering and an evaluation of the results. We put our emphasis on a generic API design to keep the implementation flexible and open for customisation.

Throughout time, as medical and epidemiological studies have grown larger in scale, the challenges associated with extracting useful and relevant information from these data has mounted. General health surveys provide a good example for such studies as they usually cover large populations and are conducted throughout long periods in multiple locations. The challenges associated with interpreting the results of such studies include: the presence of both categorical and continuous variables and the need to compare them within a single statistical framework; the presence of variations in data resulting from the technical limitations in data collection; the danger of selection and information biases in hypothesis-directed study design and implementation; and the complete inadequacy of p values in identifying significant relationships. As a solution to these challenges, we propose an end-to-end analysis workflow using the MUltivariate analysis and VISualization (MUVIS) package within R statistical software. MUVIS consists of a comprehensive set of statistical tools that follow the basic tenet of unbiased exploration of associations within a dataset. We validate its performance by applying MUVIS to data from the Yazd Health Study (YaHS). YaHS is a prospective cohort study consisting of a general health survey of more than 30 health-related measurements and a questionnaire with over 300 questions acquired from 10050 participants. Given the nature of the YaHS dataset, most of the identified associations are corroborated by a large body of medical literature. Nevertheless, some more interesting and less investigated connections were also found which are presented here. We conclude that MUVIS provides a robust statistical framework for extraction of useful and relevant information from medical datasets and their visualization in easily comprehensible ways.

The ability to catch prey is crucial for survival and reproduction and is subject to strong natural selection across predators. In many animals, prey capture demands the orchestrated activation of multiple brain regions, demonstrating the intricate interplay between sensory processing, decision-making, and motor execution. This makes prey capture a prime paradigm in neuroscience. Further, its ubiquity across species makes it ideal for comparative research and for studying the evolution of cognition. However, despite recent technological advances in the collection and analysis of behavioral data, experimental approaches for studying prey catch are lagging behind. To bridge this gap, we created PreyTouch - a novel system for performing prey capture experiments on a touch screen. PreyTouch incorporates flexible presentation of prey stimulus, accurate monitoring of predator strikes and automated rewarding. The systems real time processing enables closing the loop between predator movement and prey dynamics for studying predator-prey interactions. Further, the system is optimized for automated long-term experiments and features a web-ui for remote control and monitoring. We successfully validated PreyTouch by conducting long-term prey capture experiments on the lizard Pogona Vitticeps. The acquired data revealed the existence of prey preferences, complex prey attack patterns, and fast learning of prey dynamics. The unique properties offered by PreyTouch combined with the ubiquity of prey capture behaviors across animals establish it as a valuable platform for comparatively studying animal cognition.

SummaryThe widespread application of mass spectrometry (MS)-based proteomics in biomedical research increasingly requires robust, transparent and streamlined solutions to extract statistically reliable insights. Existing, popular tools were generally developed for specific uses in academic environments and did not fully embrace current open-source principles and best practices of software engineering. We have designed and implemented AlphaPeptStats, an inclusive python package with broad functionalities for normalization, imputation, visualization, and statistical analysis of proteomics data. It modularly builds on the established stack of Python scientific libraries, and is accompanied by a rigorous testing framework with 98% test coverage. It imports the output of a range of popular search engines. Data can be filtered and normalized according to user specifications. At its heart, AlphaPeptStats provides a wide range of robust statistical algorithms such as t-tests, ANOVA, PCA, hierarchical clustering and multiple covariate analysis - all in an automatable manner. Data visualization capabilities include heat maps, volcano plots, scatter plots in publication-ready format. AlphaPeptStats advances proteomic research through its robust tools that enable researchers to manually or automatically explore complex datasets to identify interesting patterns and outliers. AvailabilityAlphaPeptStats is implemented in Python and part of the AlphaPept framework. It is released under a permissive Apache license. The source code and one-click installers are freely available and on GitHub at https://github.com/MannLabs/alphapeptstats. Contactmmann@biochem.mpg.de, maximilian.strauss@cpr.ku.dk