CNV-seq and CNVer on RNA-seq NGS

Question

[anthonycolombo60]

This paper mentions that CNV-seq can be used on NGS data, does this mean that CNV-seq can be used on RNAseq data?

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0059128

and CNVer paper , says that the software can be used on short mated reads from NGS, does this also imply that CNVer can be used on RNAseq?

http://compbio.cs.toronto.edu/CNVer/

Thank you very much

A.Colombo

Answer 1

[mikhail.shugay]

I believe RNA-Seq is not suitable data source for CNV detection. One cannot distinguish between down-regulation of whole transcript / down-regulation of exons due to alternative splicing and expression loss due to loss of DNA copy. The original paper doesn't mention RNA-Seq either.

I think the most suitable method for CNV detection so far is array comparative hybridization (aCGH). You can extract CNVs from whole-genome sequencing also. In theory exome-seq could also provide some info on CNVs (I believe FREEC can handle it), but without any guarantees..

[anthonycolombo60]

Yes in their original paper, they explicitly say they have used aCGH. I also was under the impression that CNV can only be used on whole genome sequencing ... etc

But in the linked paper above titled "Comparative study..." on page 1 they write, " motivated by a comparative study of CNV detection methods based on aCGH technique, we conducted a comprehensive comparison of six representative CNV detection methods based on NGS under different sets of conditions."

I know that RNAseq is apart of NGS technology, so it is ambiguous as to which type of technology they are using under NGS ..

And then I get more confused with CNVer, which uses SHORT mated reads.... which I know RNAseq is comprised of.

However, I am not sure why RNAseq is not allowed, because it is not explicitly stated as disallowed. But I do know whole genome sequencing and aCGH are primarily used.

thank you very much again

AC

[mikhail.shugay]

I believe they've meant the alternatives to aCGH that come from NGS, not whole bunch of NGS methods. It is not the length of reads which defines applicability here, but the starting material, which is RNA in case of RNA-Seq :)

Please note that you should add your replies as "comments" not "answers" to your questions.