Validated Barcodes

[christophpale]

Hi,
I am interested in sequencing 40 strains of yeast ( ~12Mb)
genome size. Illumina provides 10 validated barcodes but I need
30 more. I would be grateful if people can share their custom
(and validated) barcode sequence or provide any advice
on designing our own barcodes that will not interfere with illumina
sequencing reaction. Reading through some previous posts on
multiplexing, one advice I gathered was that the first four
bases of the bar code is used by the illumina software for cluster
identification--but that leaves only 16 unique barcodes which
is insufficient.
thanks in advance
Christoph

[kshankar]

I just wanted to check and see if you did find a way to get validated barcodes more than those supplied by illumina?
thanks
Kartik

[csquared]

Illumina has 48 out now (see the tru seq small RNA kit, the other tru seq kits, including the DNA kit only have 12). Anything more than 48 and there are a number of options available. We have 144-plex per lane working routinely now, but most often use 4, 8, 12, and 48-plex using a modified version of the Illumina 3-read method.

The most important choice is if you want the 3-read method or an adaptor-based method where the first several bases of the normal sequencing read are used for the barcode.

The most important lesson we have learned working with barcodes and lots of different designs is that the use of multiplexing puts a lab's ability to make good libraries under a microscope. If you can't make pristine libraries for the samples you would like to multiplex, adding the multiplexing will add to the difficulties in the experiment.

[HESmith]

A four-base barcode gives 4^4, or 256, unique sequences. There have been a number of posts regarding barcode design. See http://seqanswers.com/forums/showthread.php?t=1369 for a clever method to validate the barcodes via checksum.

[kshankar]

Thanks you both, and to directing to the older thread very useful indeed.

I just wanted to point out this recent paper in PNAS (J Gregory Caporaso et al., doi. 10.1073/pnas.1011383107). They have utilized a 12 base error-correcting code for metagenomic applications.

KS

[Daytwa]

Originally Posted by csquared

The most important lesson we have learned working with barcodes and lots of different designs is that the use of multiplexing puts a lab's ability to make good libraries under a microscope. If you can't make pristine libraries for the samples you would like to multiplex, adding the multiplexing will add to the difficulties in the experiment.

@csquared

If possible can you please elaborate on this final point. What would be a good yardstick by which to judge whether a library is pristine? Thanks for any advice you may have!