Mixing Barcode Lengths 16S

Question

[c.v.oflynn]

Hi Guys,

For a while we have been using 7 mer barcodes, but have recently changed to 10 mers. Can anyone see any problems that using a mix of mers in one run might cause?

We do 16s Sequencing on 454, trim at 350 bp, have a modified AmpliconNoise style de-noising pipeline and use qiime for taxonomy assignment.

both 7 mers and 10 mers work fine on their own.

The 7 mers are not suffixes or prefixes of the 10 mers.

Anyone ever done this? I cant think of any obvious problems, but would love some reassurance. I am thinking possible specificity issues with higher lengths or bioinformatic hiccups ?

Thanks Guys!

Answer 1

[cts]

The only hiccup that I have encounters is if you specify a certain number of differences when matching the barcode using split_libraries.py in Qiime. In our lab we used to allow for a single mismatch in the barcode sequence but we didn't realise that some of the shorter barcodes were prefixes with one mismatch to some of the longer barcodes. That really screwed up a couple of runs before people realised what was going on.

If the shorter barcodes are all really different from the longer ones you won't have that problem though

[c.v.oflynn]

Thanks cts, that is something i'll look out for!