With the proliferation of Internet-enabled devices and the growing adoption of the Service Oriented Architecture (SOA), users increasingly require services that can meet their functional and nonfunctional requirements. In this paper, we propose a novel framework for Web service provisioning, which relies on brokered Service Level agreements (SLAs), between clients and service providers, and a publish/subscribe model to handle notifications on significant changes in quality-of-service (QoS) offerings. One of the advantages of the approach is that service providers can provide several service types and different levels of service. Moreover, the publish/subscribe model allows Service Brokers to be aware of significant changes in the QoS offerings of service providers; and consequently, be able to monitor the execution of SLAs.

In this paper, we suggest a requirement engineering process that composes (merge) use cases/scenarios and yields a formal specification of the system in form of a high-level Petri net. Use cases and scenarios are acquired in form of diagrams as defined by the Unified Modeling Language (UML). These diagrams are composed and transformed into Petri net specifications and merged to obtain a global Petri net specification capturing the behavior of the entire system. From the global specification, a system prototype can be generated and embedded in a user interface builder environment for further refinement. Based on end user feedback, the use cases and the input scenarios may be iteratively refined. The result of the overall process is a specification consisting of a global Petri net, together with the generated and refined prototype. This paper discusses some activities of this process.

FLOSYS is an interactive web-accessible bioinformatics workflow system designed to assist biologists in multi-step data analyses. FLOSYS allows the user to create complex analysis pathways (protocols) graphically, similar to drawing a flowchart: icons representing particular bioinformatics tools are dragged and dropped onto a canvas and lines connecting those icons are drawn to specify the relationships between the tools. In addition, FLOSYS permits to select input-data, execute the protocol and store the results in a personal workspace. The three-tier architecture of FLOSYS has been implemented in Java and uses a relational database system together with new technologies for distributed and web computing such as CORBA, RMI, JSP and JDBC. The prototype of FLOSYS, which is part of the bioinformatics workbench AnaBench, is accessible on-line at http://malawimonas.bcm.umontreal.ca: 8091/anabench. The entire package is available on request to academic groups who wish to have a customized local analysis environment for research or teaching.

GOBASE is a relational database containing integrated sequence, RNA secondary structure and biochemical and taxonomic information about organelles. GOBASE release 6 (summer 2002) contains over 130 000 mitochondrial sequences, an increase of 37% over the previous release, and more than 30 000 chloroplast sequences in a new auxiliary database. To handle this flood of new data, we have designed and implemented GOpop, a Java system for population and verification of the database. We have also implemented a more powerful and flexible user interface using the PHP programming language. http://megasun.bch.umontreal.ca/gobase/gobase.html.