Add GSEA as an extension

[Jari Häkkinen]

​http://www.broadinstitute.org/gsea/

GSEA source is available and more. GSEA is a GenePattern module, maybe that is a route forward?

[Peter Johansson]

I'm not a lawyer, but to me the it seems like the license of GSEA is not compatible with GPL. More serious is that one is not allowed to redistribute the software. In summary the GSEA package is an example of open, gratis, non-free software. The algorithm is well described in the in the PNAS paper, and it's straightforward to re-implement it, but it is pretty useless without the database of genesets (which is harder to reproduce). Here follows the most important portion of the license:

...
2.    LICENSE

2.1 Grant.  Subject to the terms of this Agreement, MIT hereby grants
to LICENSEE, solely for a single user and for internal research
purposes, a non-exclusive, non-transferable license to: (a) install,
execute and display the PROGRAM, only for the singe user; (b) create
BUG FIXES; and (c) modify the PROGRAM only for the purpose of making
the PROGRAM compatible with pre-existing LICENSEE networks, platforms
and applications.  Upon download of the PROGRAM, MIT also grants to
LICENSEE, solely for a single user and for internal research purposes,
a non-exclusive, non-transferable license to copy and use the
DATABASE.

2.2 No Sublicensing or Additional Rights.  In no event shall LICENSEE
sublicense or distribute the PROGRAM, modifications, BUG FIXES, or the
DATABASE without prior permission from MIT.  LICENSEE agrees not to
allow any non-employee of LICENSEE to access, view, or use the PROGRAM
or the DATABASE, unless such person is an independent contractor
performing services on behalf of LICENSEE.  LICENSEE agrees not to put
the PROGRAM or the DATABASE on a network, server, or other similar
technology that may be accessed by any individual other than the
LICENSEE.

2.3 License Limitations.  Nothing in this Agreement shall be construed
to confer any rights upon LICENSEE by implication, estoppel, or
otherwise to any computer software, trademark, intellectual property,
or patent rights of MIT, or of any other entity, except as expressly
granted herein.  LICENSEE agrees that neither the PROGRAM nor the
DATABASE shall be used as the basis of a commercial software or
hardware product and that the same shall not be rewritten in another
computer language or otherwise adapted to circumvent the need for
obtaining a license for use of the PROGRAM or the DATABASE other than
as specified by this Agreement.
...

[Jari Häkkinen]

Maybe the route forward is to create export plug-ins that generate files in a format supported by GSEA?

[Nicklas Nordborg]

This is the same license as for GenePattern and if GSEA is also a GenePattern module the GenePattern extension can maybe be used (if we decide to implement this extension). It's a maybe because different GenePattern modules requires different types of files. Some can be created by BASE plug-ins some may have to be created manually. Currently, the only file type we can create from BASE are GCT files. If I check the GSEA module on my test installation of GenePattern, it seems like it can use a GCT file but it also requires 2 or 3 additional files (gene chip database, phenotype labels, chip platform file). The output is a zip file which can be fed to the GSEALeadingEdgeViewer visualization module in GenePattern.

[Peter Johansson]

I read the README for GenePattern and understand that the idea is to not redistribute the MIT files, but rather let the user download them at installation time. Well, perhaps not the user but rather the admin. That makes me wonder how will it be possible for the admin to install the program on the BASE server without breaking:

LICENSEE agrees not to put
the PROGRAM or the DATABASE on a network, server, or other similar
technology that may be accessed by any individual other than the
LICENSEE.

[Nicklas Nordborg]

Because the GenePattern server doesn't have to be on the same server as the BASE server. BASE users that want to use the GenePattern extension must also have an account on the GenePattern server. A BASE account doesn't automatically imply access to the GenePattern server. In fact, users may install their own private GenePattern servers if the wish and use those from within BASE. In BASE you have to enter the URL, username and password for the GenePattern server that you wish to use.

But... it is almost impossible for anyone to install GenePattern on a networked computer without breaking the license. This is because the default settings will allow anyone to connect and create administrator-level accounts just by trying to login with an unused username. It is possible to lock this down, but as far as I know there is no admin-interface to create new users. So if you lock it down, you have lost the possibility to create new user accounts.