ChipSEQ on Solexa (low % align, unusable reads)

[bioinfosm]

Hi,

I thought there was some discussion, but could not find it - regarding the data output from a chipSEQ run using the Solexa GA.

There is very little dna to work with, is what the sequencing team says. But after a run through the solexa pipeline, I see under 20% passing-filter reads aligning to the reference genome. That is way too low, and that too with error rate of about 5% !

Are there any changes or workarounds as to what might be causing it? or its simply blame it on the data! The number of clusters, % pass, intensities all look fine..

[new300]

How many clusters are you getting per tile and what are the sequences that don't align, are they all the adapter sequence perhaps?

[bioinfosm]

How many clusters are you getting per tile and what are the sequences that don't align, are they all the adapter sequence perhaps?

Its about 10,000-15,000 PF clusters. I am yet to check match with adapter, is there a dataset I could use to do that? A reference sequence to check for adapter?

[bioinfosm]

And the phiX control looks perfectly fine .. less than 1% error, good data overall..

[bioinfosm]

What should be the way to detect adapter contamination in my chipSEQ data?

I tried using this as reference set for ELAND, but did not work as expected, any ideas? I would simply want to see what % of reads are adapters or adapter dimers, etc.

>Adapters1a
GATCGGAAGAGCTCGTATGCCGTCTTCTGCTTG
>Adapters1b
ACACTCTTTCCCTACACGACGCTCTTCCGATCT