Alternative index primer

[Andrew_Slatter]

Has anyone used an alternative custom indexing primer on any of the Illumina sequencing platforms ?

[suludana]

Yes, following the D.Craig paper (Nature Methods, 2008).
It works fine.

[Andrew_Slatter]

Yes, I am familiar with this paper. They are using their own index sequences within adaptors - and capturing the index sequence as part of a single read. I am interested in the more recent Illumina chemistry in which paired end reads are performed followed by a separate indexing read. As I understand it, the index read comes from the reverse complement of one of the sequencing primers, thus sequencing 'away' from the genomic sequence and into the index tag. I am interested in whether anyone has experience of using an alternative sequence for the index primer itself (not the index sequences).

[elaney_k]

Hi Andrew,

I think you're referring to the Illumina Multiplexing protocol where the index is introduced during the enrichment PCR stage of the protocol. If you contact Illumina techsupport directly they will be able to supply you with the full sequence info for these multiplexing primers. It's easy to spot where the index is and so should be easy enough to get custom-made index primers with as many different combinations as you require. As long as you only change the index part of the sequence and ensure the base composition of the index sequence is as varied as possible (i.e. mix evenly the number of A,C,G and T's for each base of the indexes in the samples that you want to pool) to aid in individual cluster identification everything should work fine. There are definitely groups that have used this method with no problems.

Elaine

[rjo]

Hi all,

Am I correct in thinking that the indexes for the llumina Multiplexing protocol are only 6 bases long? Has anyone tried hacking the protocol to get a longer index sequencing read, therefore allowing multiplexing of more samples?

[henry.wood]

Has anyone tried hacking the protocol to get a longer index sequencing read, therefore allowing multiplexing of more samples?

More? I've never been brave enough to put more than 25 samples in one lane. Are 4096 per lane not enough for you? I feel suitably humbled.

[rjo]

No need to be quite so humbled, I wasn't thinking of 4096 different MIDs. Using error correctable MIDs with e.g. a Hamming distance of 2 there are about 215 possible sequences. We regularly multiplex much higher sample numbers than this and analyse on 454.

[HESmith]

See this post for an alternative approach to validating barcodes.

[NearyJL78]

We've used an approach similar to the Craig paper, with our own 28 adapters. It worked very well, but now we are trying to Illumina multiplexing kits to speed up sample prep. This is not going as smoothly, but apparently you can multiplex up to 48 samples with the kit. Its paired-read - read 1 is the sequence, and read 2 is the adapter.