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 135

￿ the ability to execute software components written in a variety of programming

languages on geographically distributed computing resources,
access to knowledge and support for sharing. ￿

Most of the requirements listed here concern the execution phase of an e-Science
experiment. As science usually follows an iterative approach, the complexity of
experiments grows over time. Early designs are aimed at developing simple proto-
types to assess news ideas, and technologies. Once scientific application begin to
mature, the design phase becomes complex and the need for access to existing
knowledge and support for sharing is required. It then becomes important to provide
support for the design phase by allowing users within the same domain (or even across
multiple domains) to share expertise, reuse one another's tools etc. To facilitate knowl-
edge transfer either within a single scientific domain or across domains, design and
dissemination support is important. This is especially true as large projects operate
as Virtual Organizations (VO) where knowledge transfer across VOs is restricted
according to dynamic VO access policies. To facilitate the sharing of resources, a
common framework in which all scientists can perform their experiments is needed.
Throughout this chapter we use the virtual material analysis laboratory as an
example of a typical multi-physics scientific experiment. The Material Analysis of
Complex Surface (MACS) experiments attempts to identify and determine the elements
that comprise complex surfaces, regardless of the nature of the sample (Fig. 7.1 ).
The approach followed in the MACS lab experiment is generic and can be easily
applied to other application areas such as art conservation and restoration (e.g. analysis
of binding media and organic pigments in old master paintings), biomedical science
(e.g. identification of arteriosclerotic deposits in mice), and medical research
(e.g. studies of trace elements in brain tissues) (Frenkel et al. 2001 ) . Like most
scientific applications, the MACS lab experiment, as can be seen in Fig. 7.1 , consists
of three phases: preprocessing, experimentation and analysis of results.

The preprocessing phase includes a number of procedures which have to be fol-
lowed to extract the sample to be used in the experiment. To reach this goal
scientists compare various techniques and protocols and select the most appropriate
ones. Once the sample is produced, it has to be treated in order to fulfill the require-
ments of the device used in the material analysis process. It should be noted that, as
the MACS lab is currently at its first design iteration, literature currently provides
the main source of knowledge.

The experimentation process is performed with a set of specialized hardware
devices. Two devices are used in these experiments: The Fourier Transformed Infra-
Red imaging spectrometer and a 4 MeV Nuclear Microprobe. The FTIR is a non-
dispersive infrared imaging spectrometer coupled to an infrared microscope used to
examine the infrared radiation absorbed by complex surfaces. The 4 MeV Nuclear
Microprobe has a spatial resolution in the sub-micrometer range and is capable of
identifying trace elements on a high-sensitivity surface.

The analysis of results : the outcome of the experiment process is a set of data files
containing the experiment results. This data set consists of a stack of images known