The company by whom I (T. Flak) was employed needed a system to organize several types of data. These included information about the many gene targets on which we worked; the different assays that were developed for these gene targets; the different projects for clients; and gene experession information collected by TaqMan real-time PCR (see details about that sub-system on the example TaqMan Real-time PCR data system).
At the center of this system is a SQL Server database, with more than 150 tables. The schema design allowed linkage of one or more genes to each project; and one or more projects to each gene. It also incorporated linkages to public databases (e.g. Entrez Gene).
The primary client is a Windows application, written in Microsoft Visual Basic.NET (VB.NET). This application allows a user to navigate all of the data, and create new projects. It includes the TaqMan real-time PCR data, detailed here.
Between the database and the LIMS client application, we created a "data library" database interface layer. This was linked directly into the VB.NET application. In addition, specific data were exposed as an XML-SOAP Web Service; these web service functions were used by other applications, including some web-based reports, and several Excel-based tools. In addition, several Microsoft Access database used ODBC connections to the LIMS database.
We also used Spotfire DecisionSite, by means of a feature called Information Interaction Server. This allowed specialized Information Links to be built to facilitate simple data retrieval for specific purposes.
