Using a Novel Data Transformation Technique to Provide the EMBOSS Software Suite as Semantic Web Services

The EMBOSS software suite is a well established package consists of over 200 command line programs that can be strung together to perform many common tasks in Bioinformatics. To date, its users must either be proficient at the command-line for batch data processing, or they must use one of several Web-form wrappers for EMBOSS on a one-off basis. To combine the simplicity of Web-based forms with the power of the command line, we are wrapping the entire EMBOSS suite as semantic Web Services (SWS) using a novel framework of declarative rules. These rules describe how ontology-driven data in the BioMOBY SWS XML format can be converted to the various text formats EMBOSS programs use, and vice versa. The chief benefits of this approach for end-users are 1) they can chose between many BioMOBY client GUIs to create and run EMBOSS-based analysis pipelines and 2) they get automatic interoperability between EMBOSS programs and the 100's of other BioMOBY services that may be useful in answering a research question. The declarative rules approach, using XSLT and a novel "anti-XSLT" language, promotes creating modular, reusable pieces of domain technical knowledge to bridge the Web and the Semantic Web. These rule systems have successfully been applied "in the wild" to both client and server-side software, facilitating integration of bioinformatics tools and data. The framework is Open Source and is freely available from the BioMOBY code repository.