Table 1.
Examples of Meta-Resources for Computational Biology.
Figure 1.
Left panel shows the search, traversal and comparison of tools (in this case image alignment and visualization) based on their data input/output specifications.
The right panel illustrates how streaming data through independent tools (via an external graphical workflow environment, e.g., LONI Pipeline) may be facilitated by the types of data I/O parameters stored as iTools resource-specific meta-data.
Figure 2.
iTools CompBiome – the iTools Computational Biology Resourceome plug-in consists of a decentralized collection of BioSiteMaps (sitemaps of resources for biomedical computing) and a Yahoo!Search-based crawler for discovering new and updating existent BioSiteMaps anywhere on the web.
These updates propagate automatically to iTools' SandBox and are later reviewed by expert users for inclusion in the iTools DB. The distributed nature of the NCBC CompBiome may be utilized by any tool developer, user or librarian to find, compare, integrate and expand the functionality of different resources for biomedical computing. The left and right panels illustrate the XML schema definition for the BioSiteMap.xml files and the results of a manual initiation of the Yahoo!Search using the iTools CompBiome plug-in, respectively. iTools has an automated weekly crawler initiation as well as manual triggering of the crawler.
Figure 3.
The main two displays of iTools resources provide tabular (left) and graph-based (right) human interfaces to the resource database (http://iTools.ccb.ucla.edu/).
Both of these facilitate comprehensive traversal, comparison and search of resources. There are several other human and machine interfaces to the iTools database which are discussed in the text.
Figure 4.
A schematic and dynamic integration of iTools resources demonstrating interoperability of multi-disciplinary tools via graphical workflow environments.
The three nodes with dash-boundaries on the left demonstrate schematically the integration of some computational biology tools. The graphical workflow on the right depicts the practical means of using iTools meta-data to construct module descriptions and generate multidisciplinary and heterogeneous data analysis protocols.
Figure 5.
This figure illustrates the utilization of iTools for search, comparison and integration of bioinformatics tools.
In this example, we demonstrate the use of the Basic Local Alignment Search Tool (BLAST) for comparing gene and protein sequences against other nucleic sequences available in various public databases. The top row shows iTools traversal and search (keyword = blast) using the hyperbolic graphical interface, and tools comparison and investigation of interoperability using the tabular resource view panel. The bottom row shows the design of a simple BLAST analysis workflow using one specific graphical workflow environment (LONI Pipeline). This BLAST analysis protocol depicts the NCBI DB formatting, index generation and filtering using miBLAST, sequence alignment and result textual visualization.
Figure 6.
Examples of the input and output XML descriptions in the Pipeline, an integrated graphical workflow environment that mediates inter-resource communications.
If resources described in iTools include such data I/O descriptions, external interoperability environments (like the Pipeline) will be able to automatically enable construction and validation of inter-resource computational workflows.