o²S²PARC is an open, cloud-based platform for the development, execution and sharing of computational models, simulations, and data analysis pipelines, and the presentation of results. The main goal is to support the study of therapeutic neuromodulation of the autonomic nervous system. With the help of SIM-CORE, researchers can easily transform their code or models on o²S²PARC into configurable, reusable, and shareable studies, applications/workflows and services, benefiting from a large number of available modules and from a graphical user interface (GUI) that is automatically provided by o²S²PARC. The platform ensures long-term reproducibility and sustainability of computational studies and workflows.
o²S²PARC is designed to promote FAIR-ness in the computational modeling community, e.g., by enabling reuse and integration of established models and services without own development skills. All that is required to access powerful computational resources is a web-browser.
SPARC’s o²S²PARC simulation platform enabled ZMT, a Swiss-based Medtech company, to release a web-based implementation of their advanced Sim4Life desktop computational platform. S4L lite is free of charge for scientific use by students and in teaching! Read more here.
To create an o²S²PARC account, please send an email request to support. A video introduction to the platform is available here. The following resources are also available:
- Detailed user and developer documentation
- Curated and published computational models
- Information on the o²S²PARC API and Python SDK
- Platform tutorials
- Webinars on the SPARC YouTube Channel
Still more questions? Contact us here. You can also meet with the SIM-Core team during our online office hours.
Key o²S²PARC features:
Creation, integration, sharing, publication, and execution of computational models
- Modeling based on experimental data and/or biophysical principles
- Availability of more than 70 existing services as building blocks: from specialized cardiac models or compound action potential predictors to general purpose medical image segmentation and deep neural network training and application, as well as advanced physical and electrophysiological simulators
- Interoperability of services and reusable building blocks for the creation of integrated models and advanced studies
Pipelining, Guided Mode, and Applications
- Service pipelining enabled by well-defined service inputs and outputs
- Automatic handling of unit conversion and validation checks in pipelines
- Pipelines can be easily converted into step-by-step guided expert workflows or applications to
- enable non-modeling-experts to leverage sophisticated computational modeling, by hiding the computational complexity, exposing only steps requiring user interaction, and providing instructions where needed
- e.g., establish treatment planning or device safety assessment applications
Powerful functionality to simulate and study neural interfaces, neural activity, and ensuing physiological responses
- Simulation functionality for device design, as well as safety and efficacy assessment, including regulatory-grade modeling capabilities
- Anatomical models suitable for electromagnetic, acoustic or thermal exposure simulations
- A range of established neural response modeling tools, which can be coupled to the exposure modeling
- User-contributed organ physiology and response models
- Well-established tools to visualize and analyze simulation results and measurement data
- Sharing and publication of analyses, e.g., to facilitate reproducibility or enable interested parties to further explore data interactively
- Access to JupyterLab to facilitate scripted and free exploration, or Jupyter Voila for highly controlled interaction
- Presentation of results and automated creation of detailed, exportable reports
Online accessible, cloud-based, and scalable, and open
- User-friendly online resources without system requirements beyond a web-browser and without need for software installation
- Direct launch of o²S²PARC studies from the SPARC Portal, even without a user account (see Accessing Published Simulations )
- Access to scalable resources, e.g., when running large simulations or exploring vast parameter spaces for optimization
- o²S²PARC is an open source project itself; it encourages users to publish their models, services, and studies under permissive open licenses, such as MIT, but supports closed source and limited access sharing if desired.
- Findable: via the Portal and the various search functionalities of o²S²PARC
- Accessible: directly through the cloud-based platform; studies can also be shared among users and teams via a simple link or permissions
- Interoperable: nodes are self-contained and can be coupled and pipelined, and communication is ensured by the framework, obviating the need for compatible OS's, programming languages, software installations, etc.
- Reusable: computational and interactive services are modular, for maximal reusability; nodes contain code along with the runtime environment, which permits reuse of workflows, avoiding compatibility issues and ensuring sustainability and reproducibility
Studies are sustainably preserved and fully reproducible
- Run-time environments are preserved as part of computational services
- Versions are tracked and studies can be rerun using previous versions of services
Collaborative modeling, simple and flexible sharing, and citable, explorable publication
- Creation and sharing of own services, studies, and/or favorite applications
- Definition of access permissions individually or on a group level
- Facilitation of collaboration in shared studies through annotation functionality
- Curation and publication of studies, which creates a unique URL and DOI that can be cited and referenced in publications (e.g., as explorable supplementary materials) and renders the studies discoverable and accessible through the SPARC Portal.
Graphical User Interface (GUI) and Application Programing Interface (API)
- Modern and user-friendly web-interface to access o²S²PARC
- It ensures a consistent user-experience across the wide range of services provided
- User contributed services automatically benefit from a GUI provided by the platform
- API for developers and advanced users to access o²S²PARC services from within own applications and to run computational jobs using SPARC computational resources
- The API adheres to common standards, such as OpenAPI and json-schema
Quality assurance (QA)
- The platform QA comprises a large set of AI-powered code quality and security analyses, an automated unit, regression, and end-to-end testing framework, a peer-review process for code submission, as well as manual testing in staging and production environments
- Addition of meta-data on verification & validation, context-of-use, known limitations, etc. and adherence to the Ten Simple Rules for Evaluating Model Credibility – including transparent documentation thereof – is facilitated and encouraged