ChIP-chip - how much DNA do you label? equal amounts? (May/04/2010 )

Hi guys,

Just a quickie...

For my histone mod chip-chip experiments I label EQUAL amounts of DNA (ChIP and Input DNA).

A colleague of mine has just performed ChIP with an antibody for a co-activator (they want to do ChIP-seq).
Looking at the qPCR results, I would say it has not worked while they say it has (but that's another story!!).
To test the ChIP before sequencing I suggested they hybridise it to an array (much cheaper than sequencing!).
They are telling me to label 10 times less input DNA. eg: label 1000ng ChIP DNA and only 100ng input.

To me, this seems wrong. Wouldn't this just make everything enriched when you look at the array results? Isn't this just going to end up with a lot of false positives?!!

ARG. Hope someone can advise!!

Clare

I don't have any experience with ChIP-Chip but if you are basically taking a 10% input, wouldn't you simply correct by 10% when calculating % input values prior to statistical analysis? I know for ChIP-qPCR I like to use 1% input, that way my input curves are relatively close to my sample curves, minimizing any potential confounds with primer efficiency over larger ranges. Then I just correct when I do the enrichment calculations.

Perhaps with the ChIP-Chip you need to reduce your input value otherwise you may saturate your signal and no longer be in its linear range?

Just some thoughts.

MM