search

Sage-Bionetworks / fendR

Current workspace for development of drug efficacy prediction based on omic data
1 stars 0 forks source link

Determine network format #1

Closed sgosline closed 6 years ago

sgosline commented 7 years ago

We need to agree upon a fast network format to parse.

There are numerous formats listed here: http://www.home.ndexbio.org/network-formats/

The general default seems to be binary sif, which is what my network optimization programs use. But we can reformat to something else if need be.

sgosline commented 7 years ago

Current networks are in upper diagonal matrix csv format. We should put these in featherized format and also collect additional networks.

blogsdon commented 7 years ago

Birds of a feather, flock together

mjrmason commented 7 years ago

chirp chirp

On Thu, Oct 27, 2016 at 10:19 AM, Ben Logsdon notifications@github.com wrote:

Birds of a feather, flock together

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Sage-Bionetworks/Feature-Engineering-by-Network-Domain-fendR/issues/1#issuecomment-256711124, or mute the thread https://github.com/notifications/unsubscribe-auth/APXOzpWELkm5NWb4ryqmfjRm71bpj0CJks5q4N0PgaJpZM4KgdEE .

sgosline commented 7 years ago

Discussion about this focused on a comparison between feather-ized upper diagonal matrices or bigMatrix matrices. @mjrmason and @blogsdon are going to focus on this.

sgosline commented 6 years ago

Given @allaway's code and the PCSF example, we're sticking with igraph, as painful as it is!