ChIP-chip QC - quality control (Feb/17/2009 )
When doing ChIP-on-chip, after chromatin IP, but before amplification and hybridization, I heard that it's important to find out whether you did a great job in the IP, so that the samples you are comparing (CTL vs TREATMENT) has the same IP efficiency
some ppl told me I would need several positive and negative controls to normalize my samples, which is the "quality control" of my ChIP
so:
1. How many positive and negative controls are needed?
2. I'm doing experiment in mice, are there some generally accepted positive and negative controls for H3K4me3 and H3K27me3?
3. Philosophically, I don't quite understand how these controls are picked, because, say, someone studying ES cells vs MEF may know the modifications H3K36me, H2B-Ub and H3K79me are constant at position X, Y and Z, considering them good controls; but, that doesn't mean another person studying Cancer vs Normal can use these same controls.
Hi there 
We only use one locus to test if our ChIP has worked - we have done it so many times now (through to the hybridisations) that we are confident with our results. But the lab next to us does quite a few controls. So I'm not sure if there is a gold standard here.
Anyway, in our chips we don't use H3K4me3 but we do use H3K27me3. As to what controls you can use, well that does depend on your tissues. For our H3K9Ac we just use primers recommended by the supplier of the antibody at first (Gapdh). Then after a couple of chip-chips we redesigned them so that matched the highest peak for that locus.
As for H3K27me3, at first we didn't know which loci to use as a control, so we did a few chip-chips to see what was coming up (we did it in parallel with the K9 chip, assuming that the K27 worked as well!). We picked a few loci, tested them via qPCR and eventually found one So our control region encodes a protein important for brain development - which we would never expect to be expressed in our cells.
Not sure if I have been of any help, but it is early 
Clare
jiro_killua on Feb 18 2009, 02:56 AM said:
some ppl told me I would need several positive and negative controls to normalize my samples, which is the "quality control" of my ChIP
so:
1. How many positive and negative controls are needed?
2. I'm doing experiment in mice, are there some generally accepted positive and negative controls for H3K4me3 and H3K27me3?
3. Philosophically, I don't quite understand how these controls are picked, because, say, someone studying ES cells vs MEF may know the modifications H3K36me, H2B-Ub and H3K79me are constant at position X, Y and Z, considering them good controls; but, that doesn't mean another person studying Cancer vs Normal can use these same controls.
