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 144

Owner

Literature

passive data element

has image

Referenced by

Photo image

Object

Active data element

is source of
has source

processes

Sample
Extraction

relationships

Sample

has result

Sample
treatment

Result Observation

Treated-Sample

Physics-Device

Material
analysis

Conversion _tool

File Conversion

Quality
control

DC
analysis

Data Cube

Data Cube

Performed with

Analysis Tool

Certification Tool

Fig. 7.4 The PFT of the MACS lab experiments describes the flow of processes involved in data
analysis. The hatched rectangles represent either physical objects or bulk data in mass storage.
Hatched ovals represent operations, both executable and manual (Frenkel et al. 2001 )

At a later stage, less experienced users can customize a given PFT such as in the
MACS lab example, to perform a specific experiment. The user then creates an
instance of the PFT and modifies some parameters, defines input and output data
sources, attaches a new application workflow etc. PFTs enforce best practice by
providing to the domain expert a view of the flow of a given experiment. By using
PFTs the end users are assured of performing the right tasks at the right time.
However, they retain the possibility to customize the PFT to suit their own situation.
Even when not yet implemented, users should be able to combine different PFTs or
even work on subparts of a given PFT.
Because PFT and the instances created by the end users are saved in application
domain databases, they can be accessed via a simple query interface by all members
of a Virtual organization based on the view management policy (Kaletas 2004 ;
Afsarmanesh et al. 2001 ) (Fig. 7.5 ). In the current implementation the query inter-
face is limited to a search based on the PFT and the instance name; the idea being to
develop this query interface to support more sophisticated queries based on semantic
techniques. The MACS lab PFT was demonstrated at the iGrid 2002 conference
where it was applied to the analysis of a sample (paint chip) from a painting in
the FTIR spectrometer and subsequent analysis of the measurement data. Using the
PFT scientists were able to customize the original PFT to collect experimental