[Bioc-sig-seq] general quesion regarding chipseq

[Steve Lianoglou]

Hi Tobias,

On Jan 15, 2009, at 7:36 AM, Tobias Straub wrote:

> dear all
>
> i am doing my first steps into chipseq and have a few conceptual  
> questions regarding the data treatment:
>
> as a chipchip person i am used to work on ip/input ratios as  
> numerical surrogate of biological binding. now i am about to analyze  
> my first solexa datasets (have both ip and input material sequenced)  
> and it appears as if the data needs ratio calculation as well as the  
> input has quite some -probably copy dependent- heterogeneity with  
> respect to local tag reads.
>
> while ratio calculation is rather trivial in chip-chip the same is  
> not trivial -at least to me- in chipseq. is there a kind of common  
> opinion on how to perform these steps without having to manipulate  
> the raw data too much? how do you, for example, treat tag-free  
> regions in the input material?
>
> i would also like to point out that i would be interested in a  
> solution that is not based on any kind of peak detection as -in  
> fact- there are many targets that are associated with DNA in non- 
> peaking fashion.

I thought it would be relevant to put this paper on your radar, since  
it's quite recent and discusses aspects you mention (ratio calculation  
in chip-seq as well as copy-dependent heterogeneity w.r.t reads):

PeakSeq enables systematic scoring of ChIP-seq experiments relative to  
controls
http://www.nature.com/nbt/journal/v27/n1/full/nbt.1518.html

Unfortunately it still is concerned with peak calling, so I guess its  
most appropriate for ChIP for TF binding as opposed to things like  
histone modifications. Perhaps it might give you some ideas of how one  
might approach ip:input ratios and such, so you might still find it  
useful.

HTH,
-steve

--
Steve Lianoglou
Graduate Student: Physiology, Biophysics and Systems Biology
Weill Medical College of Cornell University

http://cbio.mskcc.org/~lianos