mRNA z-scores with different technologies

[Mustafa Karabulut]

Hi all,

I've been studying cBioPortal data for a while now. My aim is to understand whether there is a correlation between survival times and gene expression levels in primarily Glioblastoma patients.

I've come to an understanding of how z-scores are calculated for mRNA expression levels. As one of the members of this group has just pointed out for me, which was also clearly explained in the cBioPortal FAQ, they are all relative to the patient groups. So any mRNA z-score for the same patient in a different molecular profile can be slightly different.

However, there is something I haven't been able to figure out yet. I'm inspecting Firehose Legacy dataset for Glioblastoma.

https://www.cbioportal.org/study/summary?id=gbm_tcga

In this dataset/study, there are three different mRNA z-score profiles are provided:

* U133 (Affymetrix?) Microarray

* Agilent Microarray

* RNA Seq V2 RSEM

These three molecular profiles have different number of patients. So I would expect to see different yet comparable z-score values for the same patient. But when I dig into the data, I have observed that differences in the z-scores are larger than I expected to see. For instance, z-score for one gene jumps from 0.5 to 2.5 or vice versa for the same patient among the datasets. Since I binarize the dataset to get a view of differentially expressed genes by applying a cut-off of -2/+2, a jump from 0.5 to 2.5 is an important change.

Since the output of these three methods do not agree, I think there should be a baseline among the three that can be safely used for my analysis.

My questions:

* What is the safest baseline mRNA expression reading method if I want to use it to find a correlation between differentially expressed genes and survival time?

* Is there a measure between these mRNA reading methods that can help me understand the major differences between them?

Any comment and help is appreciated.

[debr...@mskcc.org]

Hi Mustafa,

 

Thanks for reaching out! That is an interesting exploration

 

I am no expression analysis expert but there are several studies out there comparing microarray vs RNA Seq:

 

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5955523/

 

There they mention:

 

>Decreasing running costs, higher dynamic range of expression and higher accuracy in low abundance measurements [2] are the main factors for this fast development of NGS and increasing use of RNA-Seq over microarray.

 

It makes sense then that the z scores would be different versus microarray. It also seems that it might be better to use RNASeq

 

Hope that answers your questions a bit!

 

Thanks,

Ino

--
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 on the web visit https://groups.google.com/d/msgid/cbioportal/98ba067b-88e3-4496-9dee-6bd6f09bed2bn%40googlegroups.com.

*** Only open attachments or links from trusted senders. Report phishing to inf...@mskcc.org ***

 

=====================================================================

Please note that this e-mail and any files transmitted from
Memorial Sloan Kettering Cancer Center may be privileged, confidential,
and protected from disclosure under applicable law. If the reader of
this message is not the intended recipient, or an employee or agent
responsible for delivering this message to the intended recipient,
you are hereby notified that any reading, dissemination, distribution,
copying, or other use of this communication or any of its attachments
is strictly prohibited. If you have received this communication in
error, please notify the sender immediately by replying to this message
and deleting this message, any attachments, and all copies and backups
from your computer.

[Mustafa Karabulut]

Hi Ino,

Thank you very much for taking your time looking into this and replying to me.

Your conclusion makes sense to me as well. Also, I've observed through my analysis that microarray methods output a lower number of DEGs (Differentially expressed genes) than RNASeq method does. This is one of the conclusions that the authors experimented and reported in the paper as well.

Therefore it seems that it is the best choice for me to continue the analysis by taking the RNASeq output as the base.

Sincerely,

Mustafa