Using the inSilicoDb package

With more than 500,000 genomic profiles freely available in the public domain, there is a huge amount of information accessible for computational biologists or bioinformaticians. However, accessibility to these data requires complex computational steps. Manual parsing of annotations and keywords, which is in most cases a necessary evil, tends to be time-consuming and is known to be error-prone. Also the wide variety of normalization and preprocessing methods makes the comparability of different existing studies hard, or even impossible. The inSilico initiative (http://insilico.ulb.ac.be) provides an answer to those problems with its freely available web-based database tool: the inSilico database. Starting with all public available human Affymetrix studies from Gene Expression Omnibus (GEO) [1] it provides those studies in a consistent and well curated form. With a direct connection to GenePattern [3] and the ability to export the data to different formats, the inSilico database is an efficient mean to re-analyse public datasets and improve reproducibility in genome-wide research.