"DEVELOPMENT OF A DECISION SUPPORT SYSTEM FOR BIOINFORMATICS. EXTRACTION OF PROTEIN COMPLEXES FROM A PROTEIN-PROTEIN INTERACTION NETWORK: A CASE STUDY"

Decision Support Systems and Workflow Management Systems have become essential tools for some business and scientific field. This thesis propose a new hybrid architecture for problem solving expertise and decision-making process, that aims to support high-quality research in the field of bioinformatics and system biology. The first part of the dissertation introduces the project to which belong this thesis work, i.e. the “Bioinformatics Organized Resources - an Intelligent System” (BORIS) project of the ICAR-CNR; the main goal of BORIS is to provide an helpful and effective support to researchers or experimentalist, that have no familiarity with tools and techniques to solve computational problems in bioinformatics and system biology. In the second part of the thesis, the proposed hybrid architecture is described in detail; it introduces a three-dimensional space for the BORIS system, where the viewpoints of declarative, procedural and process approaches are considered. Using the proposed architecture, the system is able to help the experimentalist choosing, for a given problem, the right tool at the right moment, to generate a navigable Workflow at different abstraction layers, extending current workflow management systems and to free the user from implementation details, assisting him in the correct configuration of algorithms/services. A case study about extraction of protein complexes from proteinprotein interaction networks is presented, in order to show how the system faces a problem and how it interacts with the user.