Using my own set of genes. I've managed to create the PPI network in using STRING, but want to know how the authors subdivided the genes by cellular component (nucleus, cytoplasm, membrane, etc.). There is no mention of it in their methods.
I'm assuming you are using the stringApp, here. In that case, there are columns that are delivered that indicate cellular location. However, as I'm sure you're aware, proteins often appear in multiple locations, so you'll need to make some decisions about which location you want a particular protein to appear. You may want to look at the Cytoscape App "localizer", which uses the String localization information to assign a single compartment for the protein.