ChIP-Seq using 100bp paired-end reads

[LostInTranslation]

Hi everyone,

I'm interested in your opinion on performing a ChIP-Seq experiment using 100bp paired-end reads on a HiSeq.
If sequencing costs is not so much an issue, would you use this setup? What would be the ideal fragment size for the library? What do you think of loosing resolution due to the longer reads?

Any helpful comment is much appreciated since I'm a complete beginner in this field.

Thanks a lot!

[Dario1984]

Why would you use paired end sequencing for ChIP-seq ? It makes sense for RNA-seq, because you might be looking for spliced reads, but not in any way for ChIP-seq.

[mudshark]

Paired end sequencing makes perfectly sense if you consider for example:

http://www.ncbi.nlm.nih.gov/pubmed/22025700

http://www.ncbi.nlm.nih.gov/pubmed/21131275

bottom line: chromatin fragment sizes are not distributed evenly along the genome. in single read sequencing expts however, this is a fundamental assumption. only with PE sequencing you can get hold of the DNA fragment length your factor is bound to.

regarding the question of resolution: it should rather get better as you know the true fragment size.

[helitron]

this completely depends on what you are looking at. If you are interested in nucleosome positioning after MNase digestion, then it perfectly makes sense (as stated in the examples above). However, for standard TF ChIPs or histone modifications it does not make a big difference since your resolution is anyway limited by the sonication range.

best