RRBS missing enz site

[sidderb]

We've just had some RRBS data returned for the first time. I am told by the CRO that it's a non-directional, Msp1 library. My head is swimming from trying to work out the combinations of read pairs etc that might be present, but one thing I thought I was sure of was that all of our reads should have the Msp1 site at the start.

However, the data for the first mate in all our samples does not appear to have the restriction site. The attached image shows the FastQC plot for "per base sequence content" for the first mate of sample 1 (left side) and second mate of sample 1 (right side). I have done nothing to this data - the CRO has adapter trimmed. We see the Msp1 site fragment in the second mate (on the right), but not for the first mate (on the left). I see the same for all the samples in this experiment.

Q1 - am I right, we expect the Msp1 site in ALL reads?
Q2 - if so, whats gone wrong here?

(am awaiting a response from the CRO)

[dpryan]

That's odd, do you also not see a spike in the GC content for the first couple bases in read #1? I just double checked some of my RRBS libraries and they all have the expected enrichment of the cut site at the beginning of both reads. Perhaps the sequencing facility trimmed off the cut site from the first read? Have you compared the read names in a couple of the files just to make sure that the read #1 files are the correct ones (weirder things have happened, though that'd be a pretty big mistake)?

[sidderb]

Just to close this off - turned out to be an issue with the CRO's sequencer and they are having to re-sequence our samples. Bit worrying that they didn't pick this up themselves...