churchmanlab / genewalk

GeneWalk identifies relevant gene functions for a biological context using network representation learning
https://churchman.med.harvard.edu/genewalk
BSD 2-Clause "Simplified" License
127 stars 14 forks source link

INDRA script generation #42

Closed xandmac closed 3 years ago

xandmac commented 3 years ago

Hello, I was looking to explore creating a custom INDRA input. I was wondering if you could provide the script used to create the INDRA network from the paper.

Thanks!

bgyori commented 3 years ago

Hi @xandmac, sorry for the slow response. First of all, here are three Python pickle files that contain INDRA statements for the experiments in the paper: https://bigmech.s3.amazonaws.com/genewalk/IsoG_HGNCidForINDRA.pkl https://bigmech.s3.amazonaws.com/genewalk/QKI_MGIforINDRA_stmts.pkl https://bigmech.s3.amazonaws.com/genewalk/JQ1_HGNCidForINDRA.pkl

In terms of building your own INDRA network, that can take additional work but it gives you flexibility in what kind of relations you want to include from multiple sources. You can see an overview of the knowledge sources that INDRA makes available here: https://github.com/sorgerlab/indra#knowledge-sources, for instance, you might process BioGRID and Signor to get Statements, save it as a pickle file, and pass that to GeneWalk as the --network_file. You can also combine this with text mining through one or more of the systems INDRA is integrated with, for instance, Reach. While this takes additional effort to set up, it allows you to derive Statements from subsets of literature that you determine as relevant, and you may relations that are not be available from pathway databases.

xandmac commented 3 years ago

Oh thanks for providing the scripts. Very helpful!

On Tue, Mar 9, 2021, 6:20 PM Robert Ietswaart notifications@github.com wrote:

Closed #42 https://github.com/churchmanlab/genewalk/issues/42.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/churchmanlab/genewalk/issues/42#event-4432341622, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKIMFDBOHK3AB4Y3O5JTSALTC2UKNANCNFSM4YB4PX5A .