The top-bar provides access to several functionalities. Clicking on the title of the application will link to its front-page. Next to it is the reload-button, which is used to force the server to recalculate a report, otherwise the report will be taken from a cache. This is useful when a report has given an error or when when it was updated somehow. Note: After a reload of the report, the browser might still try to get the file from its own cache, so you might need to use the reload button in the browser as well; I’m planning to fix this.

Next is the favourites controls, which we will discuss shortly. After that there might be a search box, if the server was started with that functionality. This search box can be configured to identify any type of entity, but generally it is only genes.

The final component of the top-bar, which I label “User”, has the user controls, which allows to login or logout, and the job bookmarks, where bookmarked actions are kept. Also this help page. Will discuss bookmarking actions latter.

As we said before, we have reports for entities, lists, and maps. Clicking the star on the favourites control makes a favourite out of what-ever we are viewing: an entity, like a study; a list of entities, such as the genes mutated in the study; or a map, which could for instance map genes to the ratio of mutations per amino-acid in the COSMIC database.

Favourites are used not only to track our interests, but as a means to communicate between different functionalities or parts of the application. We could for instance make a list of genes a favourite and then go to a study report, open the “Gene list in study” action, select our list from all the favourites and see a report of how these genes are affected in the study

The favourite controls have an option to create a new list of entities from scratch. This allows a user, for instance, to make a list with her genes of interest instead of having to browse the application to find it. Note: New lists are not made favourite be default, you still need to click on the star once the report is shown.

Actions are the primary way to integrate data and functionalities. Some actions are defined globally, but most commonly they are defined inside workflows; when a report is produced, all workflows are interrogated to see if they provide actions for it. The great majority of the actions are defined in the Genomics workflow.

When an action is selected, it loads below the action controller. An action report consists of the content of the action, and optionally, documentation a report the parameter controls and description are hidden, but they can be shown using the action control buttons in the top left of the action controller. Actions can also be opened in their own report by right/middle clicking on its button.

As with all reports, they are executed the first time they are requested and the result is cached for further access. The action control buttons can be used to reload the action, just like the reload button in the top-bar does for the complete report. The actions will of-course be recomputed for each configuration of parameters. The parameters of the action can be favourites from the user, lists or maps. These are presented to the user using an input select, and are synchronized with the user current favourites.

If an action is used often with a particular combination of parameters. For instance the “Enrichment” action of gene lists performs an over-representation based analysis of functional classes. The functional classes include several pathway databases, protein domains, gene ages, etc. If we where interested primarily in performing enrichment analysis using GO biological process terms, we can use the paperclip button to set that configuration of parameters as default, so that it always uses it initially.

Some actions may require a significant amount of time to compute. The action controller tries to reload the action periodically until it is completed (after a while it will stop insisting to not overload the server). Instead of having to wait on that page for it to finish, the user can click on the bookmark button, which will place a link in the “jobs” tab on the “Extra” menu of the top-bar, and come back to it later.

Tables in this application come enhanced with several functionalities. The rows can be sort by the values of any column.

The values of the columns can also be used to filter the rows of the table through the filter button. The filters are defined per-column, and can contain exact strings, regular expressions like /(COMPLETE|PARTIAL) RESPONSE/i, statements such as > \< == or !=. And columns containing entities they can be prefixed with the term name: to force the filtering to work over the human readible text of the entities e.g. name:/SF3B\d+/ for a column containing ‘Ensembl Gene ID’.

When a column contains entities, these are usualy listed in the table using some identifier like ‘Ensembl Gene ID’ for genes or ‘GO Term IDs’, etc. When displaying the table, it automatically substitutes these identifiers with links to their corresponding reports, and lists them by name. Additionally the column button allows the user to report the list of entities of any column or, if the first column is itselve an entity, it allows the user to report the map between the entities in this column and the values of any other column. Maps are generally used as inputs to actions and tools, but may have actions as well; gene maps have an action that performs a rank-based enrichment analysis (provided the Enrichment workflow is loaded in the server).

The content of the table can be downloaded in two forms, as TSV files, which is the original source of the table, or as Excel files. When downloding as Excel files, since are generally intended to be read by people, all entities listed that can be translated to a more human readible identifier are translated automatically. This includes change genes identifiers from Ensembl Gene ID to Associated Gene Name, or changing KEGG pathway ids to their descriptive name.