Drop the -u, -p and -n options from metaImport and validateData

[ecerami]

Here is my preferred solution to dealing with the current implementation of metaImport and validateData.

1. validateData.py and metaImport.py should be the preferred mechanisms to validate and load data, and we should introduce both early on to new people who are developing or installing cBioPortal. All other mechanisms for import, e.g. cbioportalImporter.py should be mentioned only later in the documentation. All references to old import mechanisms should be removed from all wiki pages.
2. validateData.py and metaImport.py should “just work”, e.g. out of the box and without any additional steps for two important groups of users: a) new people setting up a cBioPortal instance, who may not have a web app running; and b) cancer centers, such as DFCI and MSKCC who have password protected instances of cBioPortal.

To address 2, my preferred solution is to drop all three arguments: -u URL_SERVER | -p PORTAL_INFO_DIR | -n. To my mind, the fact that the validator requires a live web app running and/or requires that one dumps data via another script is the fundamental problem. And, we should address that first.

Assigning to @priti88 for when she is back.

[pieterlukasse]

It could be interesting to keep -u as an optional parameter for the scenario where you want to validate against a server which is not locally inatalled. I.e now it is possible to validate studies on a local desktop with only python installed thanks to -u, -n and -p options.