A key component of the SPARC Program is the SPARC Connectivity Knowledge Base of the Autonomic Nervous system, referred to as SCKAN. It is a semantic store housing a comprehensive knowledge base of autonomic nervous system (ANS) nerve to end organ connectivity. Connectivity information is derived from SPARC experts, SPARC data, literature and textbooks. SCKAN supports reasoning and offers powerful query and visualization capabilities.
SCKAN provides a central location to populate, discover, and query ANS connectivity knowledge over multiple scales. It offers an unprecedented opportunity to aggregate all information regarding ANS connectivity into a machine-readable form. Representing connectivity information using a semantic model allows us to issue queries such as, “what are the locations of neuron somas with processes that pass through spinal cord level C4?” and create a searchable visual atlas of ANS circuitry. Users of the SPARC maps will be able to query SCKAN to find more information about routes, targets and evidence. Conversely, SCKAN is used to generate SPARC maps automatically, tying a connectivity atlas to a powerful database back end.
How is SCKAN organized?
SCKAN contains statements about neuronal connectivity at the neuron population level, largely in the form of: “Neurons with somas in structure A project to structure B via nerve C.” For some connections, we have detailed connections modeled with a specification of the locations of their somas and dendrites, axon segments and synapses. Connectivity information is available as RDF triples or via SciGraph, an OWL-based graph database.
The SCKAN models connections at two levels of granularity:
A circuit represents a detailed model of ANS connectivity associated with a particular organ (e.g., bladder) or functional circuit (e.g., defensive breathing). These models are created and curated by experts in conjunction with the SPARC Anatomy Working Group (SAWG) using information from multiple sources such as data, literature, and ontologies. Each circuit contains detailed representations of neuron populations, including locations of cell bodies, dendrites, and the anatomical course of axons as they traverse anatomical structures and nerves to synapse with their targets (example). This knowledge is represented using the ApiNATOMY connectivity model, a sophisticated semantic model that represents multiscale connectivity [Kokash N and de Bono B (2021), Osanlouy M, et al. (2021) - for additional information see this video]. ApiNATOMY provides a method of chaining together axon segments to create a unified representation of tractography.
General knowledge about individual connections of CNS brainstem nuclei, ANS ganglia, nerves, and end organs:
In order to provide a comprehensive overview of the state of knowledge about ANS connectivity, the circuit-based approach is supplemented with well-known connections of ANS ganglia and nerves derived from the literature and textbooks using a Natural Language Processing (NLP) pipeline [Menke J et al. (2020) - to learn more about this process, see this video]. These types of connectivity statements do not have detailed topological information associated with them and are represented using The Neuron Phenotype Ontology (NPO), [Gillespie TH, et al. (2020)]
How do I access the connectivity within SCKAN?
SCKAN knowledge is accessible through the SPARC Connectivity FlatMaps on the SPARC Portal. With each release of the SCKAN connectivity knowledge, the maps are automatically updated.
Simple SCKAN refers to an extension of SCKAN that permits exploring connectivity knowledge in a simplified manner. It serves as a high-level abstraction layer on top of SCKAN for the sole purpose of making SCKAN's connectivity knowledge more accessible.
Simplified relationships available in Simple SCKAN express the connectivity between a neuron population and its parts are located:
- hasSomaLocation: a relation between a population of neurons and where their somas are located
- hasAxonLocation: a relation between a population of neurons and where their axons traverse
- hasDendriteLocation: a relation between a population of neurons and where their dendrites are located
- hasAxonTerminalLocation: a relation between a population of neurons and where their axon terminals (i.e., the axon presynaptic element) are located.
- hasAxonSensoryLocation: a relation between a population of neurons and the location of their sensory axon terminals (i.e., for specialized psuedounipolar neurons).
The relational properties above in Simple SCKAN serve as the 'shortcuts' for NPO's actual ontological axioms about where neuron populations reside in the PNS and brainstem. The sole purpose of this encapsulation is to allow querying and retrieving SCKAN's connectivity knowledge in a simplified, manageable manner.
Access Simple SCKAN with the SCKAN Explorer at https://services.scicrunch.io/sckan/explorer/
To use Simple SCKAN using Stardog, you first need to request a log in and password from K-Core. Here is a set of simple instructions to get you started:
Currently, direct access to the full SCKAN is limited to those who are familiar with the Cypher or SPARQL query language, we’ve created a containerized version of SCKAN that can be downloaded and installed locally along with the necessary documentation to install and query SCKAN. Find out how to access the full SCKAN here (release link) and the types of queries supported here. Updates to SCKAN are released quarterly. See the detailed Help page “How do I access SCKAN.”
Who Created SCKAN?
The SCKAN is a product of the Knowledge Management team of K-Core, a collaboration between the FAIR Data Informatics Lab at UCSD and Whitby et al, Inc. The connectivity knowledge contained in SCKAN has been contributed by a number of SPARC investigators and the SPARC Anatomy Working Group (SAWG).
To submit new queries you would like add to the SPARC Connectivity Knowledge Base, please fill out the form available here.
For additional information contact us at email@example.com