A Web - Services Approach for Integration of Heterogeneous

level. No practical data and experiments were presented to support the arguments. Jungfer et. al. (1999) reviewed the European Bioinformatics Institute (EBI)’s efforts using CORBA to integrate different bioinformatics databases. In addition to a discussion on defining IDL for the sequence data, the paper discussed several issues due to the use of the CORBA wrappers. For example, it looked at how the system states should be managed in a workflow type of query and how to map an IDL object definition to the data in a relational database. It also briefly reviewed several other prototypes built using CORBA as the basic framework. The paper shows the vast interests and the feasibility in using CORBA as the basic framework to integrate multiple bioinformatics databases. However, no quantitative data is presented to support the integration methods. There were also no discussions on communication issues over the Internet. In addition to using CORBA as the basic integration architecture, several other researchers (Siepel, et. al., 2001a; Siepel, et. al., 2001b) have used CORBA as a communication protocol to link different databases. Those researchers point out that client side integration using different tools to take advantage of different databases is also very important. Interestingly, several of these client side integration approaches used the CORBA ORB concept as the core architecture. Integrated System (ISYS) (Siepel, et. al., 2001a) is a notable example. ISYS proposed a client platform based on CORBA’s service registration concept. Each client application, such as a SequenceViewer, can register with the ISYS platform as a service and thus can be used by end users. Each server