[krobison]

Another approach is linear amplification using in vitro transcription.

http://www.ncbi.nlm.nih.gov/pubmed/21720315

Amplification-free paper

http://www.ncbi.nlm.nih.gov/pubmed/21431776

[ETHANol]

We found that about half of this unexplained dispersion is due to a bias against GC rich sequences, and that local GC content (within 10-20 bp) is the strongest determinant of the differences in sequencing coverage, and this influence decreases to about 100 bp away from the center base, where GC content matters little if at all.

How did you determine this? Did you write your own script or did you use a software package?

On GC bias and PCR, clearly the best is to go PCR-free but if you must do PCR, Kapa Biosciences says their polymerase reduces GC bias as compared to Phusion or the TruSeq polymerase (which they say is even worse then Phusion polymerase). The Kapa polymerase is more efficient (i.e. requires less cycles to achieve the same amplification) then Phusion in my hands so it would make sense.

[Arjan]

Hello. I'm working on a ChIP-Seq data set from an Illumina platform (I think GA but not sure) and I observe what looks like GA-bias in the first 5-6bp. Has anyone seen this before and does anybody have an explanation for it? So, in our data there seems to be a preference for A and G in the 5'-ends of the read.

[ETHANol]

I've seen weird biases in the first base pairs. I think it is a common problem. I just trim them off before mapping. Most mapping programs give you this option exactly for this reason. I don't run the machines so I don't know where it comes from.

[Arjan]

hmm. okay. i have been trying to find literature on this but to no success so far. thanks.

[Arjan]

also, i'm pretty sure the data i'm looking at is real and that the bias is a preference for certain fragments over others. i think chopping them off may not change my results for the good. i am however interested in an explanation for this bias.

[pmiguel]

also, i'm pretty sure the data i'm looking at is real and that the bias is a preference for certain fragments over others. i think chopping them off may not change my results for the good. i am however interested in an explanation for this bias.

I could speculate... How about: the DNA has nicks at those sites? Then sonication might preferentially snap the chromatin strands there. Or, those sequences are preferred by T4 DNA ligase and get adapter ligated at a higher efficiency than others ends.

--
Phillip

[jparsons]

hmm. okay. i have been trying to find literature on this but to no success so far. thanks.

You might try looking at http://www.ncbi.nlm.nih.gov/pubmed/20395217
The appropriate panel of figure 1 is nearly the same bias pattern as what you are showing.

[seqinganswers]

How are you fragmenting your chromatin? There is usually a bias in the first few bases if you are digesting your chromatin with micrococcal nuclease (Mnase). Here is a link to a paper that talks about this:

http://www.sciencedirect.com/science/article/pii/S0022283608014824