Peak Confidence In Findpeaks

Question

[Arturo_M]

Hello I am using Find Peaks without control data, is there any option so I can get statistical confidence on my peaks?. I see that the Min. cov. allowed is 0.001 but how do I know which is the best minimum coverage and minimum peak size ?

Thank you for your help

[KCC]

Do you have more than one replicate?

Answer 1

[apfejes]

Hi Arturo,

I answered this question in an email earlier today, and this is what I wrote to start the conversation:

Doing Chip-Seq without a control is not a very good idea precisely because you can't tell exactly what the right cutoff should be. There are several other ways that were implemented in Findpeaks, and I believe are discussed in the manual, but they are all rough approximations. The short answer is that without a control, there is no good way of knowing what the best minimum peak size you can trust should be. However, for the purpose of the experiment, you can set those to 1, and just everything so that you can decide later, if you wish.