ArrayAnalysis.org

One of the preferred ways to process data for storage in the dbNP system is via the ArrayAnalysis.org web portal. This portal features modules to perform a variety of data analysis steps, including quality control and data normalisation methods, statistical analysis, and pathway analysis.

Quality control and data normalisation modules are currently available for microarray data of several platforms (Affymetrix, Illumina) and under development for some others (Agilent, spotted arrays). Besides the raw data, the module takes a study description (file names, sample names, and experimental groups) as input. The output of the modules includes a variety of user interpretable quality control images and a table with normalised, or cleaned, data. Since this step of the analysis depends on the type of data, the specific options for output images, normalisation methods, and annotation methods, differ between platforms.

The statistics module performs limma regression modelling based statistics. The module can be used as an entry point, in which case (cleaned) data can be uploaded directly to the module. Also, it can be called from the results page of the several normalisation modules, in which case the cleaned data and the study description are automatically transferred to the module. This means that the statistics module can be easily used in a workflow-type of approach for the data types supported, but can also be used on cleaned data from other types of technology (e.g. massive parallel sequencing). The outcome of the module includes some diagnostic images and tables, and tables of the differences between experimental groups of choice, and the significances thereof.

The pathway analysis module runs pathway statistics in PathVisio, using the PathVisioRPC connection library. Also for this module, data tables can be uploaded directly to the module, or outcome of the statistics module can be transferred automatically to it. Besides an ordered list of overrepresented pathways, the module provides a hyperlinked graphical representation of the data displayed on the pathway diagrams as an output.