Summary
Objectives:
Biomedical applications, such as analysis and management of mass spectrometry proteomics
experiments, involve heterogeneous platforms and knowledge, massive data sets, and
complex algorithms. Main requirements of such applications are semantic modeling of
the experiments and data analysis, as well as high performance computational platforms.
In this paper we propose a software platform allowing to model and execute biomedical
applications on the Grid.
Methods:
Computational Grids offer the required computational power, whereas ontologies and
workflow help to face the heterogeneity of biomedical applications. In this paper
we propose the use of domain ontologies and workflow techniques for modeling biomedical
applications, whereas Grid middleware is responsible for high performance execution.
As a case study, the modeling of a proteomics experiment is discussed.
Results:
The main result is the design and first use of PROTEUS, a Grid-based problem-solving
environment for biomedical and bioinformatics applications.
Conclusion:
To manage the complexity of biomedical experiments, ontologies help to model applications
and to identify appropriate data and algorithms, workflow techniques allow to combine
the elements of such applications in a systematic way. Finally, translation of workflow
into execution plans allows the exploitation of the computational power of Grids.
Along this direction, in this paper we present PROTEUS discussing a real case study
in the proteomics domain.
Keywords
Bioinformatics - Grid-based problem-solving environments - ontology - proteomics -
workflow
