Different results from local webserver vs causalpath.org from the same input?

[wyu]

The local webserver we setup 3 days ago appears to function properly. However, the output from the same input dataset looks very different from the local vs the causalpath.org as shown in the attached. There appear to be more protein nodes on the graph, and the layouts are different as well. The same parameters are used in both cases.

Is it possible that the underlying prior databases are different? Is there a way to check on that?

Thanks for this wonderful tool. We're so close to get it up and running!

Wen

[wyu]

DGAT_test.zip

Hi, here are the files inside the analysisOut/.. folder in case you want to check out the details.

Thanks!
Wen

[ozgunbabur]

Hi Wen,

Sorry, it looks like I missed this issue. Just noticed it accidentally now.

Yes, the differences are due to updates in the priors. The webserver uses the published version of CausalPath while the latest code and resources in GitHub are updated quite a bit. We fixed some issues with PhosphoSitePlus data in Pathway Commons and added some more manually curated priors. The latest data is better.