why some ENCODE narrowPeak/broadPeak data have no strand information?

Question

[TriS]

I'm working with some ChIPSeq data from the ENCODE consortium (i.e. wgEncodeBroadHistoneK562H3k79me2StdPk.broadPeak.gz) but there is no strand information. I was reading this and they say that they use MACS (which returns strand information) and in their ChIPSeq guideline paper I couldn't find anything specific about the strand reported in narrow/broad Peak files....

The reason I'm asking is because I'd like to use Homer for motif discovery but it asks for strand information. it works even without it, but to be very precise I think I'd need those info too.

Does anybody have some input/ideas on this?

Thanks :)

Answer 1

[Emily]

I think that strand is ignored, particularly for histones, as it is misleading. Histones don't bind to one strand of the DNA, the whole of the DNA is wrapped around them twice. Yes, the reads will be of a single strand because that's how sequencing works, but they represent both strands.

[TriS]

it looks the same also for protein that bind to histones, not only for histone mods.

i.e. NCOR here (https://www.encodeproject.org/experiments/ENCSR000ATY/), the bed files still has no strand information. oki, I think I got my answer then, thanks ;)