Information Technology Reference
In-Depth Information
meta-information further to the concrete envisaged bioinformatics applica-
tions, and thus enable the actual domain-specific design of in silico experi-
ments. Accordingly, the technical contributions of this work to the application
of the novel workflow design approach to the different in silico experimenta-
tion scenarios comprise in particular:
•
The semantic enrichment of the domain models, that is, the semantic an-
notation of service interfaces and semantic classifications of services and
data types that facilitate (semi-) automatic, constraint-driven workflow
design and largely free workflow designers from dealing with service inter-
faces and data types. By lowering the required technical knowledge in this
fashion, highly heterogeneous services become accessible to researchers
who are not specially trained in programming, but who need to design
and manage complex bioinformatics analysis workflows.
•
The identification and formalization of frequently occurring workflow
constraints and domain-specific knowledge beyond the more technical se-
mantic interface descriptions in order to improve automatic workflow com-
position. Semantic service interface annotations are per se only sucient
for automatically creating generally possible service compositions, hence
often (especially when working with large domains) additional knowledge
is required for constraining the search to the actually intended workflows.
This is specifically supported by the constraint language of the loose pro-
gramming framework.
In contrast to earlier applications of the synthesis method, such as the
LaTeX process synthesis (toy) examples [98, 307] or the Telco applica-
tions [304, 297], the bioinformatics workflow scenarios as addressed in this
topic deal with an existing, historically grown, and in particular less rigidly
designed domain that comprises truly distributed services and does impose
farther reaching challenges on the workflow design method. Another central
difference is the relevance of data in the scientific applications, which de-
mands for a shared memory for flexibly exchanging data between services,
and consequently also for the integration of data-flow analysis methods into
the synthesis process in order to be able to keep track of the available data
throughout the workflow.
The four selected application examples that are discussed in this topic
cover different thematic areas (phylogenetic analyses, PCR primer design,
metabolic flux analysis and microarray data analysis), different software
components (commonly known tool collections as well as special-purpose ser-
vices), different service technologies (standard REST and SOAP Web Ser-
vices as well as specifically created jETI [202] services), and also workflows
of different complexity. In contrast to the more academic case studies car-
ried out for the conceptual study of the application domain, these projects
address specific, new application scenarios and provided experiences regard-
ing the practical applicability of the developed methods. Each example has
Search WWH ::
Custom Search