A standardized format for sequence data exchange

Abstract

At present there is no agreement upon a standard format for the presentation of sequence data; each of the major sequence databases has adopted their own format. As a result, efforts to pool these data and to develop software to manipulate the data have been hampered. A significant amount of software development time must be invested to handle the incompatibilities among these formats before software to solve biologically interesting problems can be implemented. In principle, the development of a standard format by the database distributors would be the best solution. However, because the databases have invested years of effort in the development of procedures specifically tailored to their own format, they are reluctant to change. Insisting that they convert to a new format would place an extreme burden on the already overtaxed resources of these groups. Furthermore, for certain specialized applications it is more efficient to present the data in nonstandard formats. An alternative solution is presented here. Rather than develop a single standard format for all sequence data, a standardized exchange format has been developed. This format was designed to serve as a common interface between the major formats currently in use. Data can be easily converted to and from it without significant loss of information. This alleviates difficulties inherent in dealing with multiple formats while preserving the local formats of the various databases.