Supplementary Figure 3: Anatomy of a complex Galaxy External Display Application interoperating with locally installed desktop software, which can be optionally launched over the web.

This dynamic GEDA is initially defined in the usual manner (I), however this GEDA has two sets of “dynamic_links” tagsets (II and III), where each is able to generate multiple external resource links. The first link set (II) is generated from the Galaxy built-in sites list for the Integrative Genomics Viewer (IGV), with dynamic parameters (IV) defined similarly as in (3.III). GEDA links are filtered (V) to include sites designed for directly operating on running local software or where the genome build for the dataset is enabled at the external resource. The URL (VI) is dynamically generated from a template-based parameter named “redirect_url” (X). This GEDA will work for GFF datasets (VII) and derivatives, with the URL file extension coming directly from the specific hierarchical datatype for the dataset. A (VIII) template-based parameter named “site_organism” is used to determine the genome build ID valid for the IGV software from a look-up table or as directly specified, with the “strip” attribute indicating to remove any whitespace surrounding the populated parameter value. A (IX) template-based parameter named “jnlp” dynamically generates a Java Network Launch Protocol (JNLP) description file that can be used launch a new instance of the IGV desktop application with the user’s GFF dataset. The (X) redirect_url used for the external resource URL (VI) is created in one of three different versions, depending upon the type of external resource site definition (local, web, or jnlp). The second set of “dynamic_links” (III) is defined from a Galaxy Data Table that is populated from an external URL (http://igv.broadinstitute.org/genomes/genomes.txt), enabling automatic syncing with the latest IGV genome build support. A filter (XI) is applied to ensure that only datasets with genome builds available within this IGV database will have GEDA links created. A URL (XII) is created that forwards the user and dataset content URL directly to the launch IGV page at the Broad Institute, along with a defined track name and genome build. The GFF dataset content (XIII) is defined as in (VII). Reference: Thorvaldsdóttir H, Robinson JT, Mesirov JP. Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform 2013;14:178–92. https://doi.org/10.1093/bib/bbs017.