An increasing number of biological databases have been made available in the form of RDF that are accessible through SPARQL endpoints. These endpoints o er a valuable opportunity to utilize the RDF datasets as integrative databases. However, writing a SPARQL query often becomes a burden for biologists; thus, an easy-to-use querying tool for the SPARQL endpoints is necessary. Here, we developed SPANG, a command-line SPARQL client supporting query generation. SPANG can dynamically generate typical SPARQL queries according to the commandline arguments. SPANG can also use SPARQL templates existing in local system or on the Web. Further, SPANG allows a user to combine multiple queries, each with distinct target endpoints, via Unix pipe. These features enable easy access to RDF datasets through SPARQL endpoints, and enhance the integrative analysis of various biological data distributed across the Web. SPANG is freely available from http://purl.org/net/spang.
Because of the rapid progress in biotechnologies, various types of biological data have been rapidly accumulating. The Semantic Web technology has attracted attention as a promising approach for integrating such growing heterogeneous data; thus, an increasing number of biological databases have been made available in the form of RDF through SPARQL endpoints [1{4]. Although these SPARQL endpoints provide an opportunity to utilize the RDF datasets as integrative databases, using SPARQL language often requires cumbersome coding tasks, such as pre x declaration of URIs and writing common code patterns repeatedly. Automation of such troublesome tasks will help the researchers use the biological databases through SPARQL endpoints. Here, we developed a novel SPARQL client, SPANG, for reducing the burden of coding SPARQL and increasing the reusability of written SPARQL codes.
SPANG is a command-line client that can dynamically generate SPARQL queries according to the command-line arguments. Basically, a single SPANG command submits a query to an endpoint. SPANG has two modes of operations: (i) SPARQL shortcut mode, where typical query patterns are generated according to command-line options; and (ii) SPARQL template mode, where
SPANG: SPARQL client supporting query generation the speci ed SPARQL template and parameters in the command line are used to generate a runtime query. The speci ed SPARQL template can be either a local le or a le on the Web. A normal SPARQL query that does not include parameters can also be executed in the SPARQL template mode. SPANG has several other mechanisms to simplify the cumbersome tasks in SPARQL querying; (a) pre x declarations described in con guration les are used for runtime autocompletion; (b) nicknames for SPARQL endpoints de ned in con guration les can be used in the command line; and (c) SPARQL template libraries in the local system can be looked up by specifying a template name in the command line. Although the distributed SPANG package provides prede ned con guration les and a template library for general use, each user can extend the settings by user-de ned les.
Whereas a single SPANG command can submits a query to a speci c
endpoint, SPANG also enables combinatorial execution of multiple queries, each
with distinct target endpoint, by connecting respective SPANG processes via
Unix pipe. This functionality is similar to that of the SPARQL 1.1 federated
query using the SERVICE keyword [
Thus, SPANG enables easy access to RDF datasets through SPARQL
endpoints, and also facilitate combinatorial use of distributed databases across the
Web. These functionalities will enhance the integrative analysis of various
biological data toward knowledge discovery. As a practical application, we will show a
set of example queries using the UniProt SPARQL endpoint [