Semantic MOBY 1 of 5 Schiltz, Gessler, Stein Semantic MOBY Position Paper for the W3C Workshop on Semantic Web for Life Sciences 27-28 October 2004, Cambridge, Massachusetts USA Gary Schiltz <gss@ncgr.org >, Damian Gessler <ddg@ncgr.org > National Center for Genome Resources, Santa Fe, New Mexico USA Lincoln Stein <lstein@cshl.edu > Cold Spring Harbor Laboratory, Cold Spring Harbor, New York USA Introduction This paper describes Semantic MOBY (www.semanticmoby.org ), a Semantic Web-based architecture and open source software project for integrating web-hosted resources in the field of bioinformatics. During its design and development, a number of questions were addressed regarding standards and technologies related to the Semantic Web. The authors hope that lessons learned (discussed in this paper and in [1]) while addressing these questions will be valuable to the W3C and the broader Semantic Web community as they guide the evolution of the Semantic Web. The MOBY Project The name MOBY is derived from the acronym MOBY-DIC: Model Organism Bring Your own - Database Interconnectivity Conference. This series of conferences, first held in 2001, was for scientists and engineers working in bioinformatics – a field concerned with the management and analysis of biological data, particularly in genomics and proteomics. Participants sought to formulate approaches to make their respective databases and analysis services widely available to the community, and to make it easier to integrate these resources. Because a large number of bioinformatics data sources and analysis services are available over the Web, integration efforts including MOBY mainly focus on Web-based data and services. This paper uses the term "integration" to refer to the description, discovery, and engagement of bioinformatics resources. The integration needs of two groups of users, namely bioinformaticians and biologists, are addressed by Semantic MOBY. Bioinformaticians focus on the development and use of software tools for analysis of biological data. They require tool sets to facilitate development of bioinformatics software, particularly applications that engage multiple distributed resources and merge results. For example, an application could perform a BLAST search to find proteins that contain similar amino acid sequences, followed by a literature search for articles that discuss the function of those proteins in connection with a particular disease. Conversely, biologists are primarily concerned with applying their expertise in biology, and even those with significant computer skills may not have programming experience. What is more important is to discover and engage resources provided by bioinformaticians, and if possible, to do so without the necessity of installing or maintaining software. Software aimed at biologists is usually browser-based. As a result of the first conference, a set of use cases and scenarios [2] were developed to guide the effort. Two complementary approaches to integration were proposed [3]. The first approach, based on the Web Services paradigm, is referred to as "MOBY Services;" that approach is not described here. The second approach, based on concepts from the Semantic Web and dubbed "Semantic MOBY," is the subject of this position paper. Information about the MOBY effort in general (referred to generically as "BioMOBY"), as well as MOBY Services and Semantic MOBY, may be found at www.biomoby.org . Although Semantic MOBY and MOBY Services are specifically targeted at the field of bioinformatics (based on the expertise of the investigators and the agencies that sponsored the project), both architectures are quite general and are suitable for integration efforts in other fields. Shortcomings of Traditional Architectures A number of technologies have been used to provide interoperability in the context of distributed software development. A few examples include CORBA [4], Enterprise JavaBeans [5], and Web Services [6], as well as ad hoc CGI. Although these technologies are quite different, the Semantic MOBY team identified three seemingly ubiquitous