ArrayExpress is an international public repository for well-annotated microarray data, including gene expression, comparative genomic hybridization (CGH) and ChIP on Chip
microarray experiments. ArrayExpress Data Warehouse which is one of the four components of ArrayExpress, contains curator-selected subset of MIAME compliant data from ArrayExpress that are re-annotated for compatibility and includes updated information about the genes represented on microarrays. There are two ways of querying the data warehouse provided at the moment. The default (“simple”) interface provides users with a fixed query path through the warehouse and is easy to use, but not very flexible. The “advanced” interface allows doing anything, but is very hard to use; it is probably more useful as a developers’ tool, providing some level of abstraction above SQL level direct database access. The main goal of this thesis project is investigating ways to provide better ArrayExpress data warehouse query functionality to end users, developing a working system and demonstrating integration into the ArrayExpress infrastructure. The idea is to provide an interface where users can enter free text like searching in Google instead of the current simple query form where users have to enter gene(s) and species information. One can type anything in the description field of the interface and the system would understand what the user wants, so one could type in 'yeast time series' and get all time series experiments on S. cervisiae and S. pombe, or one could type in a much more complex statement that would include customized restrictions like “experiment type: time series AND species: yeast” etc. In this project we developed a search engine which provides the above mentioned flexibilities to the users. Also an intuitive interface for advanced querying is proposed at the end of this report. |