This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Normalizing RPPA data counts from TCGA

Hello all,

For a study, we decided to look into the expression levels of a particular gene in TCGA data, following which we were intrigued to look into the protein expression levels (RPPA) as well.

The aforementioned data are all available on cBioPortal. My question is, what is method followed to normalize the RPPA data.

Thanks in advance.

Edit: Protein expression values seem to be NaN for all tumor types checked. Has anyone obtained protein expression data from cBioPortal?

tcga protein expression rppa
• 4,255 views • link
updated by Kevin Blighe Kevin Blighe • written by vinayjrao vinayjrao

1 answer

In the TCGA project, not all proteins were profiled via RPPA. So, if your protein of interest was not profiled, then there will be no data for it.

Most likely, considering that you have obtained the data from cBioportal, the data is already normalised / standardised to the Z-scale, in which case you do not have to do any further processing. Check the exact name of the data-type that you select - it should have 'Z-scores' somewhere in the title.

Kevin

• 0 views • link
written by Kevin Blighe Kevin Blighe

Hi Kevin,

Thanks for the reply. If I don't consider the expression values with Z-scores, is it still acceptable?

Also, I downloaded RSEM values (gene expression). Is it OK to plot the data as the absolute RSEM values, or log2(RSEM)?

Thanks again.

• 0 views • link
written by vinayjrao vinayjrao

Hey, you are now referring to another [RNA-seq] dataset? What was the name of the exact data that you obtained from the cBioPortal page? If it is RSEM RNA-seq, then you can input that to DESeq2 for normalisation: Transcript abundance files and tximport input.

• 0 views • link
written by Kevin Blighe Kevin Blighe

Yeah, so we wanted to analyze RAN and protein expression for a couple genes. The RNA expression data were RSEM values. I also read that RSEM values are already normalized, so the question is to plot the absolute values of the log.

• 0 views • link
written by vinayjrao vinayjrao

Can you show the exact data processing steps that you have performed?

• 0 views • link
written by Kevin Blighe Kevin Blighe

Hi,

I'm sorry for the delay.

After download the RSEM values, I merged them in order of the tumor type using shell. Without any further normalization, this data was imported to R and plotted.

• 0 views • link
written by vinayjrao vinayjrao

Log in to answer this question.

More posts like this