Data Analysis Reporting and Visualization

To provide meaningful drug-discovery information, the raw data from a mass spectrometer must typically go through a series of processing steps. Data processing of mass spectrometer data can have very different characteristics and be highly dependent on the application. For example, vendor software packages that perform quantitative analysis of specific MRM transistions based on peak integration and calibration regression are well established. In the absence of a LIMS system, a manual process typically would involve being exported to a text-based file. The text file could be opened with Microsoft Excel and is manually reformatted, sorted, and filtered. This type of activity would be a good candidate for automation with Visual Basic for Applications (see the following paragraphs).

To interface postacquisition data into databases requires the ability to query the database and provide exactly the data that are needed into a specified report format. Often, even with commercial LIMS applications, the provided reports do not provide enough information or place it in the correct format. In this case, there are a number of approaches that can be used to create custom queries. Many LIMS vendors are providing integrated custom reporting tools based on commercially available query engines such as Crystal Reports.

Structured Query Language (SQL) is an industry-standard language for interaction with databases. A number of powerful commands and functions are available; however, an analyst may only need to read data for reports, and therefore only needs to use the SELECT command. Ad hoc database query tools, such as BRIO Intelligence Designer (Sunnyvale, California), can allow end users with no in-depth knowledge of SQL to rapidly build and execute database queries through a graphical interface. This tool is also an excellent way for those new to database queries to learn SQL, or extract lengthy SQL commands for insertion into other applications. In addition, it is possible to create query reports that span multiple database systems. This approach can be particularly useful in drug-discovery studies where automated reporting is required for streamlining the data release to clients.

Typically, for BRIOTM application development a database administrator can provide a catalog file that contains the necessary connection information for the database of interest. Permission must also typically be granted for access to the database outside of the normal data-access application. BRIO files can be deployed as Web applications provided the client machine has the BRIO plug-in for Microsoft Internet Explorer. The graphing options within BRIO are somewhat limited for complex scientific charting and visualization. However, ad hoc queries have the flexibility and power to create custom data pivot tables that can be automatically exported

Spotfire (Somerville, Massachusetts) is a powerful data-visualization tool that can accept data from a number of different data sources. Spotfire has gained in popularity in high throughput screening areas and is capable of integrating with chemical-structure databases. Examples of Spotfire for visualization of experimental ADME, in silico, and chemical data have been reported [61].

Was this article helpful?

0 0

Post a comment