Suggestion for Oncoprint Sort By Options

34 views
Skip to first unread message

Amy Makler

unread,
Feb 13, 2017, 10:59:11 AM2/13/17
to cbiop...@googlegroups.com
Good morning,

Cbioportal is one of my go to tools because of just how much data is available. I have a suggestion for the OncoPrint data. Would it be possible to add a co-occurrence type of sorting? I know the data is available through the “Mutual exclusivity” tab, but it would be great to have this kind of clustering visual through OncoPrint, without having to manually move all the lines around.

Thank you and best regards,
Amy Makler

JJ Gao

unread,
Feb 17, 2017, 8:51:13 AM2/17/17
to ama...@fau.edu, cbiop...@googlegroups.com
Dear Amy,

Thank you for your suggestion. Could you please give us an example of "co-occurrence type of sorting"?

Thanks,
-JJ  


--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+unsubscribe@googlegroups.com.
To post to this group, send email to cbiop...@googlegroups.com.
Visit this group at https://groups.google.com/group/cbioportal.
For more options, visit https://groups.google.com/d/optout.

Amy Makler

unread,
Feb 17, 2017, 10:10:48 AM2/17/17
to JJ Gao, cbiop...@googlegroups.com


Dear JJ,

For example if Gene X and Gene Y have significant co-occurrence, then in the Oncoprint section, under the Sort dropdown, maybe have something there for sorting by patient and CNA, to cluster those patients who have significant co-occurence together. In the attached screen shot, the same patient shows CNA in FASLG, MIR206, ABCB1, and BCL2. BCL2 would have to be moved manually to complete the clustering. With smaller data, moving the rows around isn't too bad, but with larger data, it can become time consuming. We also do not know which genes have significant co-occurrence until we look at the mutual exclusivity tab. For this example, only one pair of genes, BCL2 and FASLG, has significant co-occurrence when we look under the exclusivity tab.

With a sort by option allowing those significantly co-occurring genes to be clustered together in Oncoprint, it would provide a very nice visual to accompany the mutual exclusivity table. Perhaps it may be difficult to cluster together on Oncoprint, but introducing some type of visualization would be of great help.

Thank you,
Amy Makler




Inline image 1
Inline image 2

JJ Gao

unread,
Feb 17, 2017, 2:30:58 PM2/17/17
to Amy Makler, cbiop...@googlegroups.com
Hi Amy,

Thank you for the detailed example. We will discuss and see if we can support sorting / clustering genes.

Best,
-JJ
Reply all
Reply to author
Forward
0 new messages