Hello,
I've been wondering what barcodes to use for some 454 amplicon sequencing: A couple of new publications confidently suggest that their barcodes are better and than the barcodes from Hamady et al. 2008 (which is referenced over 200 times) published in nature protocols. The newer articles say that the "Hamady" barcodes aren't really error-correcting. See krishnan et al. 2012 and bystrykh 2012, particularly krishnan et al. 2012 which provides new barcodes in their supplmentary. The coding theory they use to make their suggested barcodes is pretty lost on me, but I think I get the point. What I do not know is how much of a problem using the popular Hamady barcodes would be in practice (probably not much).
Also,
also wondering about the possible bias introduced by barcodes. Turns out there are a couple of interesting papers on the subject, particularly Berry et al. 2011. They find that there is bias introduced by different barcodes, and suggest a remedy. Specifically, they amplify sample with template specific primer. Then, take 1uL of that reaction and perform another 5 cycle PCR with a primer that has the 454 adapter, barcode, and template specific primer. They show that this can increase similarity between replicates. It is certainly worth consideration since we will often be comparing samples very similar to one another. My guess is that it might not matter so much considering we usually see at least 2% change between communites (by bray-curtis).
Any thoughts on either of these things?
Thanks,
David
I've been wondering what barcodes to use for some 454 amplicon sequencing: A couple of new publications confidently suggest that their barcodes are better and than the barcodes from Hamady et al. 2008 (which is referenced over 200 times) published in nature protocols. The newer articles say that the "Hamady" barcodes aren't really error-correcting. See krishnan et al. 2012 and bystrykh 2012, particularly krishnan et al. 2012 which provides new barcodes in their supplmentary. The coding theory they use to make their suggested barcodes is pretty lost on me, but I think I get the point. What I do not know is how much of a problem using the popular Hamady barcodes would be in practice (probably not much).
Also,
also wondering about the possible bias introduced by barcodes. Turns out there are a couple of interesting papers on the subject, particularly Berry et al. 2011. They find that there is bias introduced by different barcodes, and suggest a remedy. Specifically, they amplify sample with template specific primer. Then, take 1uL of that reaction and perform another 5 cycle PCR with a primer that has the 454 adapter, barcode, and template specific primer. They show that this can increase similarity between replicates. It is certainly worth consideration since we will often be comparing samples very similar to one another. My guess is that it might not matter so much considering we usually see at least 2% change between communites (by bray-curtis).
Any thoughts on either of these things?
Thanks,
David