Starting concentration of sample/control in ChIP Seq

[mnandita]

Given the little amounts of DNA obtained after ChIP, are analysis algorithms equipped to normalize data across input/mock/test ips if different amounts of starting material were used in each case to make the libraries?

So if i have 10 ng of IP test, but 40 ng of input and 100 ng of mock, should I necessarily make all libraries from 10 ng only, or can I take all the DNA I have for each of the experiments?

[ETHANol]

You should absolutely use the smallest amount of DNA from your samples in all your libraries. In your example that would be 10 ng. The main reason is because PCR induces biases with each cycle (exponentially I believe), so if you are amplifying one sample less then other, the PCR bias won't be transferred equally between them.

More simply, samples should be treated equally, to do that you need to use the same amount of starting material.

One other thing, mock IP isn't really a good control. Just use input and your ChIP.

[mnandita]

Thanks Ethan. I agree with that (reg. starting with same amounts).

As for the mock control- if your Ab is not very good, and there is sufficient amount of non-specific DNA pull down that does not show up in the input, how do you control for it without a mock? I do not think the mock is dispensable, except in cases where you have a very specific Ab.