Dose CNV analysis need the number of samples exceed a therehold?

Question

[zhaolianhe1992]

Hi all,

I checked some CNV analysis software including cnvnator, cnvseq, xhmm, cnvkit, exomeCNV etc.

Many or most of them needs sample number exceed 20 or 10 or a certain number

Say if I only have one sample as case and one sample as control, because CNV is very eazy to be affected by PCR,or read depth,or some other noises. it is very difficult to get a proper result with just 2 samples?

Thank you!

lianhe

[zhaolianhe1992]

if there is a way to deal with such situation,just 2 sample?

Answer 1

[Irsan]

Assuming you want to perform copy number profiling based on whole exome sequencing data, you should use the off-target reads, not the on-target reads. See here for a relevant biostars post. If you use the off-target reads, you don't have to deal with the systematic biases between baits and therefore don't need multiple samples to correct for these systematic biases.