Development of Standard Services for Integrating Rodent Brain Atlases Based on Spatial Location

Ilya Zaslavsky (San Diego Supercomputer Center, University of California San Diego, USA), Asif Memon (University of California San Diego, USA), David Little (University of California San Diego, USA), David Valentine (San Diego Supercomputer Center, University of California San Diego, USA), Stephen Larson (University of California San Diego, USA), Raphael Ritz (INCF, Sweden), Steven Lamont (University of California San Diego, USA), Richard Baldock (MRC Human Genetics Unit and IGMM, UK), Rembrandt Bakker (Radboud University, Netherlands), Albert Burger (MRC Human Genetics Unit and Heriot-Watt University, UK), Christian Haselgrove (University of Massachusetts Medical Center, USA), Mike Hawrylycz (Allen Institute for Brain Science, USA), Andreas Hess (Institute of Experimental and Clinical Pharmacology and Toxicology, Germany), Lydia Ng (Allen Institute for Brain Science, USA), Daren Lee (University of California Los Angeles, USA), Seth Ruffins (Biological Imaging Center, California Institute of Technology, USA), Fons Verbeek (Leiden Institute of Advanced Computer Science, Netherlands), Janis Breeze (INCF, Sweden), Jyl Boline (Informed Minds, USA)

The goal of the INCF Digital Atlasing Infrastructure (DAI) Task Force is to develop scientific and technical foundations and to support implementation of a distributed system of brain atlas resources that can be published, shared and integrated based on spatial characteristics of the information. These spatial characteristics can include different ways of referencing location in the brain (by coordinates, by anatomic labels, by spatial placement rules, verbally), different types of spatial relationships (common topological and metric relationships, as well as specific spatial predicates used by neuroscientists) and different types of spatial arrangement and patterns. One of the key principles of this development is reliance on international information model and service interface standards for spatial data. The standards that form the backbone of the infrastructure include:

·         Waxholm Markup Language (WaxML): XML schema expressing a uniform information model for key elements that form a brain atlas: coordinate reference systems, transformations, landmarks, points of interest (POI), images, labels, annotations and other objects returned on POI-based requests. WaxML borrows spatial object descriptions from Geography Markup Language (GML), an international standard for spatial data encoding (ISO 19136.)

·         Atlas Web Services: service interfaces for querying and updating atlas data, which return information as WaxML-encoded documents. The atlas services follow the Open Geospatial Consortium's Web Processing Service (WPS) interface standard, which provides a standard framework for describing, invoking and chaining web requests, specifically oriented to spatial data processing functions.

The INCF-DAI Task Force defined the following atlas service requests using the above services:

A.    Core atlas service requests

·         Service capability descriptions: GetCapabilities and DescribeProcess. These requests, mandated by the WPS standard, provide descriptions of functions (processes) included in an atlas service.

·         Descriptions of spatial reference systems (SRS) hosted by an atlas service implemented at an atlas hub: ListSRSs, DescribeSRS. These requests return coordinate system origin, units, definitions of coordinate axes and other SRS metadata formatted as WaxML documents. Currently, descriptions of the following coordinate systems are included: the INCF Waxholm Space (WHS) pre-release and release versions, the reference atlas, 3D volume and AGEA of the Allen Brain Atlas project, and the Paxinos-Franklin stereotaxic coordinates for the mouse brain. The functions are implemented at all atlas hubs that publish data in a coordinate system other than WHS.

·         Spatial transformations: ListTransformations, DescribeTransformation, TransformPOI. These functions support description of execution of coordinate transformations between the known spatial reference systems, in particular transformations between the canonical WHS space and other SRS. Additional coordinate systems and transformations are added to the system as new images and volumes are registered, to support spatial integration of atlas data. The latter function accepts coordinates of a point of interest (POI), in any coordinate system declared within INCF-DAI, and generates coordinates of the POI in a target atlas space.

·         Structure retrieval: GetStructureNamesByPOI. This method supports structure lookup for a canonical set of segmentations defined for each atlas, returning WaxML descriptions of structures found in the vicinity of a POI, where the POI can be formulated in any known coordinate system.

B.    Atlas service requests that are not mandatory but are likely to be implemented at one or several atlas hubs. These include such POI-based requests as Get2DImagesByPOI; GetCorrelationMapByPOI; GetGenesByPOI, GetAnnotationsByPOI, which accept a point of interest in any known spatial reference system and return a respective WaxML document from a given atlas service.

The atlas services are being implemented for five hubs: Allen Brain Atlas hub, UCSD Cell-Centered Database hub, Edinburgh Mouse Atlas Project hub, a WHS hub, and a central INCF atlas hub. The service are accessed from several brain atlas client applications, and registered in a neuroscience data portal. Using this functionality of INCF-DAI, neuroscientists can discover how information from different atlases (imagery, gene expression, labels, annotations, etc.) is related by spatial location in the brain, integrate data from multiple atlas sources, and incorporate spatial information processing in their analysis and visualization applications.

Preferred presentation format: Demo
Topic: Digital atlasing

Document Actions