An informal off-site informatics meeting was held on November 15, 1994, as an adjunct to the DOE Human Genome Program Contractor-Grantee Workshop in Santa Fe, New Mexico. The discussion at the National Center for Genome Resources involved about 25 NIH and DOE investigators. Its purpose was to develop ideas and a working plan for an interoperable framework of genome-analysis tools. An earlier discussion on this topic was organized by David States at the Cold Spring Harbor Genome Mapping and Sequencing meeting in May 1994 [see HGN 6(4), 16 (November 1994)]. As in the earlier session, the group strongly agreed about the desirability of a framework to provide researchers with seamless and transparent access to an array of sequence-analysis tools and databases.
The Santa Fe meeting centered around a plan to develop a multiple-client environment that could access a variety of Internet servers. It would facilitate genome analysis and data retrieval, entry, and annotation, transparently passing data types and control among the various tools. In this approach, individual tool designers would attach their servers to the framework by submitting a client code and information about their server and input-output data types to a central library. The modular, flexible framework would allow growth and new types of analysis and data and facilitate community-wide interoperability.
Several goals were reached at the meeting:
* Agreement that semantic issues could be overcome and would not prevent framework construction, a divisive issue at previous sessions.
* An understanding that large centers often require tbatch modet processing, whereas individual users most often want interactive analysis. Both should be incorporated by design.
* Discussion of ASN.1 and other systems for standard format description.
* Consideration of data-type or object-exchange mechanisms such as CORBA.
* Identification of an initial suite of tools to be included in a prototype framework.
Participants also agreed on the importance of building and testing a prototype for evaluating methods and comparing approaches. With this in mind, several groups are beginning to work on systems that would incorporate tools to fetch entries from Genome Sequence Data Base, GenBank (Trademark), and perhaps the Genome Data Base; access sequence-analysis systems such as XGRAIL; provide links to such sequence-comparison tools as BLAST, genQuest, and Biopoet; and facilitate data entry and annotation.
The prototype would demonstrate several major project goals, including the ability to navigate and access distributed Internet tools within a single environment, exchange data among tools, and use interoperable systems for data retrieval, analysis, and annotation. The group hoped the prototype would be available by May, when members will reconvene at the Cold Spring Harbor meeting. |
| Author: Aaron Hall |
| |
Author Bio:
|
| This article can be searched using: social sciences, health colorado at denver & health sciences, 10 social sciences |
|
 |
|
