PMC

(1) User requests to view scheduled session. (2) Meta Server replies with session object. (3) User copies Data Server Path string into Uploader application. (4) Uploader application sends scientific data to Data Server. (5) Data Server stores scientific data. (6) Person requests to view downloading interface. (7) Meta Server replies with downloading interface. (8) User issues download request. (9) Meta Server forwards download request to Data Server. (10) Data Server sends scientific data to Meta Server. (11) Meta Server forwards scientific data to Internet PC. (12) User stores scientific data on Internet PC.

The communication process of how a client successfully initiates a flow to the server.

When requesting information from a partner server, the local server establishes a connection to the remote server and checks if this connection is available (see link 1). On start-up, The client application (EVidenceFINDER) establishes a connection to the local server (see link 2), receives connection data from its local server and then checks connection to the remote server (link 3). If the connection succeeds, the name of the remote server is displayed on the user interface. When the user selects the remote server's name, the connection to the local server (link 2) is closed and the connection to the remote server (link 3) is established instead.

The general workflow of the GraPES server. The user can lookup precalculated MaGS information or, alternately, run their own novel predictions on the web server. The GRaPES system is separated into a Processing Server and Client Server. The user's browser interacts with the Client Server to request results and submit jobs under their session identifier. The Client Server then forwards these requests alongside an authentication token to the Processing Server via HTTP calls. The Processing Server responds to requests for 1) results by retrieving them from the associated database or cache, 2) queue information by considering the state of pending and completed jobs, and 3) job submissions by storing the necessary information for future processing. MaGSeq jobs are scheduled to Job Runners in approximate order of submission while attempting to prevent single users from monopolizing the queue. Proteomic background distributions and their associated kernel density estimates of protein features are precalculated and cached in memory.