Questions about some terminologies.

Question

[mangfu100]

Hi all.

I have simple questions for understanding copy number variations.

As many reported studied have shown that there are extensive variability across exons.

and with this concept, I have been reading below paragraphs, but I can't catch the reason.

Here are paragraph I didn't understand:

Similarly array comparative genomic hybridization (aCGH) methodologies, the ratio of read count between a test and a reference sample is usually preferred to a single sample analysis in order to control for the typically extensive variability in capture efficiency across exons.

How can we control variability across exons just using ratio between test set and ref set?

I didn't get it.

Answer 1

[Devon Ryan]

The problem with calling CNVs in single-samples is that there's a lot of variability in things like mappability and capture efficiency across the genome. These alone can create both false-positives and false-negatives. To get around this, if we process a control sample in exactly the same way, then we can "simply" compare coverage in our test sample to it and then we have a baseline that's (mostly) pre-corrected for these problems, since it should be affected in the same locations to the same extent.

[mangfu100]

I got your answer. Thanks.

So I would think in the same way that in case of point mutations, the control will be a normal samples while test is a tumor samples to make some baseline to detect variations.

[Devon Ryan]

If you're interested in somatic mutations, then yes that's correct (the overlap in calls between the tumour and normal samples from the same individual would then be germline mutations...assuming they aren't just inherited variants).

[mangfu100]

very much appreciate your answer :)

Actually I used published tools (ExomeDepth) to detect CNVs for my research. and it was surprised to see in this paper that they use many control samples to reinforce their reference sets.

Incredible..