Users and Groups

Users in OpenCGA

Any study has intrinsically defined three different membership levels or user groups:

Owner

The owner of an study is the user that creates the study. Because of that an study will always have one and only one owner, who will be able to perform any action over the data contained in the study. There are two actions that are only possible for the owner user: delete the study and assign or remove users to/from admins groups (see next section).

Administrative groups

OpenCGA defines two reserved groups that will have some special behaviour.

Admins

Every Study in OpenCGA contains a special group called admins. This group will contain a list of users that will be able to do most of the administrative work the owner might want other users to do. Users belonging to this group will be able to perform almost any action except for the two ones that are only allowed for the owner of the study. Special operations that only these users will be able to perform are create/update/delete groups of users, create/update/delete variable sets and assign/remove permissions to other users/groups.

Members

Apart from admins, there is also an special group called members. Any user with any kind of granted access to the study will automatically belong to this group. The main aim of this group is to keep track of the users with any access to the study, but it also has other advantages such as:

  • The admin users might want to predefined some permissions any member of a study will have. In such a case, admin users will just add new users to that group and those users will automatically be granted the permissions the group has.

  • If an admin user wants to completely revoke any permission to one user, by removing that user from the members group, OpenCGA will automatically search for any permissions set for that user in any entity and remove it.

Decision Algorithm

The next schema provides a visual explanation of the algorithm implemented in Catalog for deciding whether the user has or not access to the data in the context of a study.

There are two circumstances under which the algorithm behaves as follows:

  • If the user and any of the groups where the user belongs to have permissions defined for one entry, the permissions that will be actually used will be the user's.

  • In case the user belongs to more than one group and those groups are assigned different permissions for one concrete entry, the effective permissions that will be used will be the union of the permissions found in all those groups.

Last updated