Overview

What Is the Cell Centered Database?

The Cell Centered Database (CCDB), launched in 2002 as one of the first Internet databases for cellular imaging data, makes 3D microscopic imaging data available to the structural biology and neuroscience communities. This database houses structural and protein distribution information derived from confocal, multiphoton, and electron microscopy, including correlated microscopy. Through this publicly accessible resource based at the University of California, San Diego (UCSD), researchers can query the rich and unique datasets derived from electron tomography and explore new algorithms for segmenting and visualizing data.

What Kind of Data is Stored in the CCDB?

The CCDB was designed to store and manage cell level information from tissue, cultured cells, and subcellular fractions, regardless of the type of tissue being studied. Because a significant portion of the tomographic work performed at National Center for Microscopy and Imaging Research (NCMIR) concerns such neuronal structures as spiny dendrites and synaptic structures, the CCDB contains several specializations specific for neuronal data. For example, morphological information on the structure of individual nerve cells is included from cells filled with fluorescent intracellular dyes (see Bushong et al., 2002 for details). Filled neurons are stored as series of optical slices, both with and without deconvolution, and also as branching tree structures traced using Neurolucida (Microbrightfield Corp., Colchester, VT, USA).

The CCDB models the entire process of reconstruction, from specimen preparation to segmentation and analysis. [View schema] A volume reconstruction is stored along with pointers to all of the raw images and the processing details required to reconstruct the volume from the raw data. Each object segmented from the 3D volume is stored as a separate object indexed to the parent reconstruction. Four types of segmented objects are modeled in the CCDB:

surface objects: polygonal surface meshes representing 3D objects in the reconstruction, extracted using either isosurfacing methods or manual contouring
contour objects: a series of contours that have not been fitted with a surface
volume objects: subvolumes containing an object of interest
tree objects: skeletons of branching objects like dendrites and axons, derived from Neurolucida (Microbrightfield, Inc., VT)

Each object is stored with such measurements as surface area, volume, length, number, and labeling intensity. Whenever possible, parsers are written for the output of analysis programs to allow results to be uploaded directly into the CCDB. For example, measurement summaries for tree objects are uploaded directly from the output of NeuroExplorer, an analysis program for Neurolucida derived data.

The CCDB is actively developing novel data types that allow abstract representations of the segmented objects described above to be stored directly in the database. Through the use of these data types, users will be able to query properties of the segmented objects directly, rather than descriptive data stored with the segmented object.

Who Created the CCDB and Why?

The National Center for Microscopy and Imaging Research (NCMIR) created the CCDB in March 2002 to accommodate high-resolution 3D light and electron microscopic reconstructions. The CCDB also is one of a series of linked and federated databases across multiple biological scales as part of the Biomedical Informatics Research Network (BIRN) project, a National Institutes of Health (NIH) initiative that uses emerging cyberinfrastructure, such as high-speed networks, distributed high-performance computing, and software and data integration capabilities, to foster large-scale biomedical science collaborations.

The goals of the CCDB project are to:

investigate methods for creating databases for multiscale imaging data
provide a means to share high resolution 3D microscopic imaging information from light and electron microscopy
create image management systems for large, complex microscopy data
develop means to link cell level data to data acquired at different scales an modalities through data federation tools

Affiliates

The CCDB is affiliated with NCMIR, the National Biomedical Computation Resource (NBCR), BIRN, and the Center for Research in Biological Systems (CRBS), a UCSD-organized research unit that exists to provide human resources, high-technology equipment, and administrative services to scientists researching cell structure and function relationships in various system processes.

NCMIR develops technologies to bridge understanding of biological systems between the gross anatomical and molecular scales and to make these technologies broadly available to biomedical researchers. NCMIR provides expertise, infrastructure, technological development, and an environment in which new information about the 3D ultrastructure of tissues, cells, and macromolecular complexes may be accurately and easily obtained and analyzed. Visit NCMIR at www.ncmir.ucsd.edu.

Data and File Management

CCDB's robust datasets are managed using an object-relational framework implemented in Oracle 9i. For more information about the CCDB's data and file management system, click here.

Enhancements

The CCDB is under active development. In the coming months, the CCDB will continue to populate the database and further enhance its functionality and usability. Future plans include the development of more user-friendly advanced query and data input forms that will maximize the benefits offered by the CCDB's rich data model. A feature that outputs data in XML format will also be added. Updates are posted regularly.

Future updates include:

Advanced Query System. Users will be able to query data, properties, and advanced data types.

Protein Distribution Analysis. New algorithms and pre-computed spatial histograms will allow users to analyze protein density after submitting a query to images in the CCDB.

The Smart Atlas. This knowledge-based query system utilizes geographical information system (GIS) technology to manipulate spatial brain data. Users will be able to query data in the CCDB through the Smart Atlas database and retrieve and integrate multiscale data on mouse models of human disease registered to a particular location in the brain. For more information, visit the BIRN site. Read more about the Smart Atlas on page 6 of the BIRN Newsletter. [PDF file]