The largest overlap between the in vivo and in vitro responses is found for inflammatory cytokines and chemokines. Within the common up-regulated set of 383 genes, several immunological processes are represented by a substantial number of genes. These can be summarized in order of decreasing overlap with the in vitro studies. Combining gene expression data from multiple BEZ235 distributor studies creates the possibility to compare effects and look for common or specific responses. In this study, we focused on in vivo acute lung inflammation models. We included allergic asthma models and exposures to air pollutants, as these also cause pulmonary inflammation and therefore provide gene expression data to which the nature and the extent of infection responses can be compared. When data from different studies are combined, it should be kept in mind that not al studies are equally comparable, as there are differences between inflammation models as well as between species, time points, as well as other practical details on how the study was performed. However, combining studies also results in a larger data set, which allows for an analysis to reveal additional information that would not be apparent in the original studies used. When more microarray data on pulmonary inflammation models will become available in the future, it can therefore be expected that the number of identifiable common and specific responsive genes and pathways will increase. When different studies employ different methods in analyzing raw data this can cause unwanted differences on the normalized data. For this reason we used the same normalization procedure on all raw two-color array data. Downloaded Affymetrix data were already normalized according to standard methods. Also, as the included studies used several kinds of microarrays, the initially collected data contain a large number of genes for which only data from one or two studies are available. Therefore, to reduce the influence of missing data on the analysis, we also applied a filtering on the set of included genes, as described in the Methods section. The criteria were chosen so that a sufficiently large number of genes was included and small adjustments to the criteria had only a minor effect on the resulting SP600125 JNK inhibitor clustering. The data used contained information for 45 compared exposures that could be grouped into five main categories, namely chemical, bacterial, viral, parasitic, and allergic asthma models. These data led us to identify a common cluster of 383 genes with a similar in vivo response pattern characterizing acute lung inflammation. Of these 383 genes, 120 were previously identified as belonging to an in vitro common infection response. Within this cluster there were subsets associated with more specific functional roles such as the response to bacterial and viral infection, cytokine and chemokine signaling, general inflammatory response, and the response to parasites and allergic asthma models.