The EU nowadays witnesses increasing demands with regard to chemical safety. In particular, animal-based test models need to be replaced preferably by robust, non-animal assays in vitro/in silico which better predict human toxicity in vivo, are less costly, and are socially better acceptable. Consumer's and patient's health will benefit and competitiveness of EU's chemical manufacturing industry will be increased. For developing such assays, FP6/FP7 Research Programmes are exploiting the revenues of data-dense genomics technologies. However, till date, there is no infrastructure foreseen which aims at capturing all data produced by toxicogenomics (TGX) projects, in a standardized, harmonized and sustainable manner. Data may thus evaporate. The lack of such an infrastructure also prevents innovative breakthroughs from meta-analyses of joint databases and systems modeling.Driven by these needs of the TGX research community, diXa will focus on networking activities, for building a web-based, openly accessible and sustainable e-infrastructure for capturing TGX data, and for linking this to available data bases holding chemico/physico/toxicological information, and to data bases on molecular medicine, thus crossing traditional borders between scientific disciplines and reaching out to other research communities. To advance data sharing, diXa will ensure clear communication channels with and deliver commonly agreed core service support to the TGX research community, by providing SOPs for seamless data sharing, and by offering quality assessments and newly developed tools and techniques for data management, all supported by hands-on training. Through its joint research initiative, by using data available from its data infrastructure, diXa will demonstrate the feasibility of its approach by performing cross-platform integrative statistical analyses, and cross-study meta-analyses, to create a systems model for predicting chemical-induced liver injury.