Currently, the TP53 Mutation Database (Olivier et al, 2002) of the International Agency for the Research on Cancer (IARC) is the biggest and most detailed database on mutations of the TP53 human gene. Release 9 includes 19,809 somatic mutations whose descriptions have been derived from 1,769 papers, 1,732 of which are included in the Medline database. Vocabularies and standardized annotations, like the ICD-O International Classification of Diseases - Oncology and SMOMED, are used for data input.
Queries of the database can be executed on-line and imply a human interaction. Results can be downloaded in a tab delimited format, but partial donwloads, e.g., subsets of all information available for a given mutation, are not allowed. Main accessibility limitations are lacking of links to bioinformatics tools and human-orientation of the interface. In the perspective of an integrated bioinformatics network environment, interoperability between services should be stressed as much as possible, both by setting up links to external resources and by implementing machine-oriented interfaces.
Our main objective is the improvement of machine-oriented accessibility of the IARC TP53 database. This goal can be achieved, at least partially, by implementing a new SRS based service and by designing and exploiting an ad hoc Document Type Definition.
Within this site, the TP53 IARC somatic mutations and literature references databases are available on-line and can be easily and effectively queried by using the usual SRS query forms. Due to the careful definition of SRS data fields, terms included in the controlled vocabularies that were used during data input can also be used from within the SRS extended query form, thus allowing for a data-driven search.
Moreover, defined SRS links between the two databases allow to retrieve data
from one database by also imposing restrictions on the other.
As an example, consider:
HTML links from the literature references database to PubMed site at NCBI
allow a direct access to available information, either abstract or full text.
Purpose views (MutationData, SampleData and PersonInfo) can be used to
focus attention on related information of a datasets,
hiding not pertinent information, thus achieving a more compact output.
By using SRS views manager, personal data views can also be created.
Finally, users, after a fruitful query,
can opt for displaying and/or downloading results as either a text file,
a generic XML file (an XML file based on the database description in the
SRS configuration files) or a specific XML file (an XML file based on the
newly defined DTD).
Bibliography