adapter multimers

[misscleo]

Hi,
I am trying to multiplex on the GAII with homemade PE adapters (incl 5nt barcodes) and all I see after cloning my libraries are adapter multimers. Each fragment contains the sequences for hybridization to the flowcell, so PCR is working (Im using Phusion), but my insert DNA is nowhere to be found. I am following the Illumina Protocol with some modifications suggested by Craig et al and Quail et al (sonication, double size selection, SPRI PCR cleanup etc). I designed my adapters to be identical to the Illumina PE adapters with the barcode tacked on the 3' end and a T overhang. I havent modified them in any special way, just HPLC purified. Any suggestions?

[richard]

Hi, I had a similar problem. Try not to used the superclean ligase recommended in the Sanger paper. Good luck!

[seqgirl123]

I would do a sanity check following the same protocol again, but using the Illumina PE adapter on your material. If they produce adapter multimers too, you rule out that it is indeed not your home made adapters but another step.

[caddymob]

This is always a problem. In my experience the best way to get away from the adapters is to always do your gel cut above at least 300bp. I have done ligations of just the adapter (both the illumina manufactured adapter and our own IDT synthesized adapters with the phosphorothioate mod as detailed by Quail et al) -- and they will concatemerize up to 200-250bp. And since they are the most abundant sequence in the library, you will see them far more than your actual target...

Richard -- why do you say NOT to use ultra pure ligase? I've tried both and they seem the same to me, and in the end the ultrapure from enzymatics is actually cheaper than NEB...

[misscleo]

Thanks for the tips everyone, esp caddymob.

I recently heard that incorporating LNAs towards the 3' end of the adapters can help facilitate formation of the Y structure. Has anyone tried that?

I will try the phosphorothioate mod and size selecting closer to 600bp.

[richard]

I indeed agree that size selection is a good way to avoid dimers in your library. For fully double stranded oligos however I found that the ultrapure ligase gave fare more multimers that the Promega T4 DNA ligase. I can't explain, but that is what my experiment showed.