Use ChIP-seq input to calculate copy number?

[Jeannine]

Hi,

I was wondering if anyone knows if I can use the reads from a ChIP input to estimate copy number of a specific locus?
I'd like to calculate the rDNA and mtDNA copy number, however only have ChIP-seq input reads and not whole genome seq reads...

Any help is highly appreciated!
Thanks,
Jeannine

[Brian Bushnell]

I don't see how absolute quantification would be possible, given the extreme bias of ChIP reads. Maybe you could get a rough estimate of relative abundance between two members of the same species for a specific locus, if you are able to determine that the variance of normalized peak heights between individuals.

[Jeannine]

I'm not talking about ChIP reads but input reads. The only bias in this case would be the cross linking and shearing, and low coverage (2x) as the input sample is normally not sequenced at great depth.

[dpryan]

Depending on exactly how the input reads were made they're likely to show some bias. Even illumina reads in general don't show uniform coverage, which is why CNV detection methods typically use a sample vs. control paradigm. That combined with your low overall coverage will make it near impossible to accurately call CNVs with that dataset.

[Jeannine]

Hm, alright.. I actually already tried estimating the copy number using the input data and got funny results. At least I know now that the cells aren't the problem but the dataset!
Thanks!