Support for Cooperative Experiments in e-Science: From Scientific Work flows to Knowledge Sharing - Identification of Ligand Binding Site and Protein-Protein Interaction Area - page 134

applications from six scientific domains have been considered: food informatics,
medical diagnosis and imaging, biodiversity research, bioinformatics, high energy
physics, and telescience.
The goal shared by all the applications developed within the VL-e project is to
take advantage of recent achievements in building large scale computing infrastruc-
tures and information systems. Regardless of the scientific domain, in terms of
computing and information management similar requirements can be identified for
applications such as developing models for predicting late-year bird migration
volumes (for the purposes of ensuring airspace safety) (van Belle et al. 2007 ) ,
visualizing high resolution correlated multi-spectral images (Broersen et al. 2007 ) ,
or developing interactive visualization tools for fused functional magnetic resonance
imaging (Blaas et al. 2007 ). However, it is more challenging to identify, given such
a large collaboration, common characteristics in term of methods, techniques and
tools and to abstract support for these features and requirements into a shared frame-
work, which avoids redundancy in performing similar tasks across different e-Science
domains by promoting exchange of resources.
The rest of the chapter is organized as follows: Section 7.2 presents a typical
application use case which is used throughout the chapter to map the concepts intro-
duced to a concrete example. Section 7.3 describes the different phases composing
the lifecycle of a complex e-Science experiment. Section 7.4 presents a survey of
the state of art in the field of workflow management systems. This survey focuses
upon three main points: design, execution, and dissemination and sharing. Section 7.5
describes an approach to constructing an e-Science framework: it describes in detail
the main components and tools developed to achieve this vision; specifically the
Process Flow Template (PFT) to describe the logic of the experiment, ontology-
based tools (OWT) to automate the generation of the PFT data structure, the
workflow management system (WS-VLAM) to execute work fl ows on geographically
distributed computing and storage resources, a bus-like architecture (Workflow
Bus) to allow the design of a meta-workflow composed of an application created in
multiple workflow management systems, and, finally, a service to optimize data
sharing across workflows composed from web services.

7.2 Motivation - A Typical e-Science Application

Scientific experimentation often involve applications which are data-intensive,
CPU- intensive. Moreover, some applications may require access to special devices.
They usually have similar requirements concerning the use of computing resources
and the ability to execute legacy or third-party applications, to perform parameter
analysis (parameter sweep), or the automation of repetitive tasks (job farming).
Typical e-Science applications have a set of common requirements:

￿ on-demand access to computing resources through a uniform interface,

￿ on-demand access to storage resources,