[Biodevelopers] xml database for bioinformatics

[John Merrells]

Joseph Landman wrote:

>Much of the XML DB's out there simply serialize the structure.  This
>may work to a degree, but it ignores the power in RDBMS, and makes the
>XML -> RDBMS mapping (and inverse mapping) slow.  Could you provide more
>details on what your system does?  My interest is in part because I have
>built an XML -> RDBMS mapper which doesn't serialize the data structure,
>nor ties one to a particular database.  
>
At present we store the document in its native form... as a single
blob of text. The document collection can then be configured to
maintain various kinds of indexes, which the query processor
uses to resolve queries. This provides for excellent performance
of selection of whole documents from large document sets.

Does your scheme map XML onto a generic or specific relational
schema? By which I mean does the user have to define a relational
schema for their XML and then provide a mapping between the
two... or do you have a generic schema... like a table where each
row is a node like this...

{ node_id, node_type, node_name, value, parent_node, previous_node, 
next_node }

Where might I find some a sample data, and even better a sample
set of queries?

John