Task creation and running (GUI). Here is shown a phylogenomic inference task on human Notch1 protein. The graph associated to the phylogenomic pipeline is displayed on the left part of the figure, as a graphical tree. We introduced a virtual concept of "step of work" that allows to show a cyclic oriented graph as a tree. At each step one or several units can act. (e.g.: at the step named "Protein", the unit "sequenceProvider", whose role is to read protein sequences from a file, will work). At the level just next to the current unit, are represented the units that will be activated as its continuation. (e.g. "BLAST" unit follows "sequenceProvider" cause the first treatment executed on a protein is the BLAST search). At the same graphical level as nodes related to a unit, are shown the parameters which can be customized for this unit (e.g. on "sequenceProvider" unit, the parameter "taxeid" (the query sequence's taxon) or parameter "$filePath" (path to the file with proteins to be analyzed)). The task given as an example in the figure was currently running when we took the screenshot. In green are shown units that finished their work, in red those which are running, in blue those which are not running. One can guess, by observing buttons on the right part of the figure, that the presented task: is an instance of pipeline model named "__ProtPhyloGenix__" (the one which produces phylogenomic inference studies for proteins), can be interrupted at any time, can be cloned (when user want to run it again modifying only several parameters), and finally explored through the scientific results web pages already produced according to the execution.