absolute counts Vs alteration frequency

6 views
Skip to first unread message

Sadaf Abedi

unread,
Jun 30, 2025, 10:23:18 AM6/30/25
to cBioPortal for Cancer Genomics Discussion Group
Hey there,

I am trying to figure out my gene "Fam43a" upregulation or downregulation in various cancer types (TCGA pancancer atlas studies) and in the "cancer type summery'' tab, in the Y-axis when I choose absolute counts, the type of cancers that shows to have highest expressions totally change from the graph with alteration frequency. I am very new to bioinformatic and I do not understand difference of theses 2 words so much.

JJ Gao

unread,
Jun 30, 2025, 11:39:48 AM6/30/25
to Sadaf Abedi, cBioPortal for Cancer Genomics Discussion Group
Hi Sadaf,

Thanks for contacting us.

The z-scores and related concepts are explained here: https://docs.cbioportal.org/user-guide/faq/#what-are-mrna-and-microrna-z-scores

When you query your gene, it will
1. use the zscore cutoff (by default 2) to calculate up/down-regulated (meaning zscores higher/lower than the 2 or -2 by default) for each sample to defined whether a sample is altered in the gene (as you see in OncoPrint)
2. compute the alteration frequency per cancer type. 
3. when you switch to absolute count, the y-axis will be the number of samples instead of frequency.

Note: you might find the Plots tab helpful when investigating gene expression data.

I hope this helps. Please feel free to let us know if you have additional questions.

Best,
-JJ

On Mon, Jun 30, 2025 at 10:23 AM Sadaf Abedi <sadaa...@gmail.com> wrote:
Hey there,

I am trying to figure out my gene "Fam43a" upregulation or downregulation in various cancer types (TCGA pancancer atlas studies) and in the "cancer type summery'' tab, in the Y-axis when I choose absolute counts, the type of cancers that shows to have highest expressions totally change from the graph with alteration frequency. I am very new to bioinformatic and I do not understand difference of theses 2 words so much.

--
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+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cbioportal/4ec31785-65ef-4bb3-b136-80e28c2c41e3n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages