Replies: 1 comment 1 reply
-
Hello, I was wondering if I could get any further information on the source of this part of your data? I see you've stated GO consortium as a data source, but we're having trouble finding them, could you confirm if you used a specific .gaf file from there to identify the protein-GO term links represented in node-label or if it was retrieved from another source? That would be very helpful. |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
When building the node-label matrix in the ogbn-proteins dataset, where did you retrieve the protein annotations to the labels/GO-terms from? I cannot seem to find a sizeable part of them in the GO .gaf files, even after converting (most of) the protein IDs from String to Uniprot. In case you did use a .gaf file, if you could also point out the version/date of the file used I'd be much obliged.
Beta Was this translation helpful? Give feedback.
All reactions