Is it necessary to perform paired-end sequencing for ChIP-Seq

[aquleaf]

Hi all. I'm trying to do ChIP-Seq in two different fragmentation ways including sonication and MNase treatment, respectively. I'm wondering is it necessary to perform paired-end sequencing for ChIP-Seq? Although paired-end sequencing can help to distinguish the PCR duplicates from the duplicates in the original unamplified DNA, the possibility to observe the duplicates in the original DNA sample should be low with sonication. Right? I'm not sure about MNase treatment.

Thanks very much! Any reply will be greatly appreciated.

Best

[Ohad]

In my humble opinion, it is not mandatory. Though I guess it depends on what protein you precipitate.

Paired-end gives you the complete fragment location and therefore, you can adjust your coverage to the middle of the fragment where your protein is probably located. However, this information can also be drawn from cross-correlation analysis of single-end sequencing.

Regarding duplicates, it's true that pair-end sequencing can mark duplicates more accurate, and I don't know if MNase treatment yields more of them, but I know that there is no one answer for this issue. I remember a publication claiming less than 1% of the data is actual PCR duplications. Duplicates number can also be affected by the type of protein I think... I would assume that transcription factor will show more duplications than histone modification. Also, I never perform the library preparation myself, but I wonder what if your protein is located at the linker DNA, should MNase be used at all?

To conclude, I tend to say that pair-end sequencing is much more important when performing RNA-seq where intron were removed. But I don't know the cost difference and if it's "not a big deal" then why not?

Hope this helps