ddRAD - barcodes vs. indices

[ewars]

Hi all,

I'm planning on running ddRAD on 1 plate of 96 individuals. I want to pool all of the individuals into one lane of HiSeq (I've done the calculations to figure out coverage etc, and it's fine because my organism has a small genome).

My question is this: why does everyone use 48 barcoded adapters and 2 indices for their 96 individuals? The adapters are much more expensive than barcodes. Is there a reason why I couldn't/shouldn't use, say, 8 barcoded adapters and 12 indices?

[nucacidhunter]

That technically is possible and cost effective option if one has limited number of samples. Using more barcoded adapter reduces hand on time by requiring less size selection, PCR reactions and quantification for pooling before sequencing. It also increases chance of obtaining equal number of reads for samples.

[SNPsaurus]

Does it also have to do with getting sufficient diversity of nucleotides in the barcodes? Although it is possible to design 8 barcodes with even nucleotide representation and with the new software update diversity is less of an issue.

[ewars]

Thanks nucacidhunter and SNPsaurus. After looking through the protocols carefully I see your points.