Computational models generated from public data. Different approaches highlight genes and associations with potential influential roles in response to injury. A. Top network modules identified with the ClusterONE algorithm . Network nodes shown as rectangles: genes with exclusive membership in the modules shown, diamonds: genes with multiple module membership. Edges represent co-expression associations. B. Top candidate regulatory circuit inferred with the RegNet algorithm . Edges represent gene co-expression association, with arrows defining the direction of the association between ptgis and the other genes according to linear regression models. The latter are shown for each gene-gene association. C. Snapshot of network module identified with the WGCNA algorithm, which contains genes involved in A and B, including ptgis (higher resolution image in Additional file 2). D. List of genes with significant concordant expression values between our model derivation dataset  and a dataset obtained from an independent study based on amputation model . Between-dataset correlation values are shown.