For efficient organization of research in the domain of bioinformatics it is required to organize properly the relevant information in specific research areas. One of the important outcomes of such organization would be provision of access to and querying of a large number of distributed information sources including various data on the primary and spatial structure of DNA and RNA macromolecules, proteins and their complexes as well as data on peculiarities of their interactions with each other.
To provide for semantic integration of nonsystematic population of autonomous information sources kept by different information providers into a well-structured information collection it is required to create the global unified representation of the existing information sources and services. To reach that it is proposed to form a special middleware consisting of the subject mediators. For each subject mediator, the application domain model is to be defined by the experts in the field. This model may include specifications of data structures, terminologies (thesauri), concepts (ontologies), methods applicable to data, processes (workflows), characteristic for the domain. These definitions constitute specification of a subject mediator. Due to that, mediators provide a uniform query interface to the multiple data and procedure service sources, thereby freeing the users from having to locate the relevant sources, query each one in isolation, and combine manually the information from them.