Index-driven XML data integration to support functional Genomics

We identify a new type of data integration problem that arises in functional genomics research in the context of large-scale experiments involving arrays, 2-dimensional protein gels and mass-spectrometry. We explore the current practice of data analysis that involves repeated web queries iterating over long lists of gene or protein names. We postulate a new approach to solve this problem, applicable to data sets stored in XML format. We propose to discover data redundancies using an XML index we construct and to remove them from the results returned by the query. We combine XML indexing with queries carried out on top of relational tables. We believe our approach could support semi-automated data integration such as that required in the interpretation of large-scale biological experiments.