A novel tool to compare microarray data from different experiments

Abstract

In recent years there is a transition from experiments involving a small number of conditions, with an emphasis on specific genes induced or repressed, to experiments involving hundreds of conditions with patterns of global gene expression. This represents enormous amounts of information, and to compare data obtained in different experiments bioinformatic tools are required. Microarrays, for example, enable the transcript levels of an entire genome to be measured simultaneously, and have become important for crop species where yet little genome information is available. In these cases, comparative análisis of microarray results requires additional hard work. In this context, a Perl postprocessing tool was developed to compare several datasets obtained with the same chip platform. It was originally created to analyze results obtained from microarray experiments using a cDNA microarrays platform (MartinezGodoy M.A. et al. 2008. BMC.Genomics 9, 318) of citrus plants exposed to diverse biotic stresses. The result for each assay, alter preprocessing raw data with another software such as MapMan, is a tabdelimited text file with a list of spot or EST identifiers (ID) and their associated parameters. Optionally, their respective Arabidopsis thaliana orthologous genes can be assigned and functionally grouped, and finally differentially expressed genes are selected. The script searches for genes requested by the user, and allows to compare them between different experiments. The input file could be a citrus ID list or an Arabidopsis identifier (ATG) list, and the user is allowed to choose which parameters will be returned. The output file contains the ID/ATG, the selected parameters for each experiment, and may incluye annotations available for each spot if the user selected it. To aid in the implementation, particularly for biologists, a browser-based interface provides easier interaction with users. This interface was written Perl using CGI library. It could be deployed either to use local sites or shared bioinformatics resources. Considering results obtained by other oomics techniques or platforms, simple modifications of this script will enhance its applicability, to allow comparison among experiments performed on different platforms. The increasing number of pathogen genomes present great opportunities for the development of new diagnostics.