Can I run these as "Small RNA" on MiSeq?

[sweetph3]

I have a collaborator who amplified V3 regions using the following primers:
Forward

aatgatacggcgaccaccgagatctacactctttccctacacgacgctcttccgatctNNNNCCTACGGGAGGCAGCAG

Reverse (one of 10)

caagcagaagacggcatacgagatgtgactggagttcagacgtgtgctcttccgatctACATCGATTACCGCGGCTGCTGG

Italics are flow cell adapter sequence, underline is Illumina sequencing primer site, bold caps is index sequence, and plain caps is V3 primer. I know the index site is not where Illumina usually puts it (these are downstream of the sequencing primer, Illumina puts them upstream of the sequencing primer), but he swears he used these before with the index site moved like this and it worked.

The indexes used do not match normal "TruSeq" indexes, but they do match the TruSeq Small RNA primers (RPI1, RPI2, etc.).

My questions:
1. Can the MiSeq read the indexes and demultiplex if the indexes are moved?
2. If the indexes can be read, can I run these samples as "Small RNA" instead of "TruSeq" on the MiSeq so I can plug in the matching index IDs? Or do I leave the demultiplexing up to my collaborator who says he had to do it before anyway so can do it again.
3. The sequencing primer site for "Small RNA" is different than the sequencing primer site for "TruSeq"? Will this confuse the sequencer?

Thank you for your input!

[kcchan]

The sequencing read won't work at all under default configuration because it's not compatible with Illumina's index primer. The way it's designed right now you'll actually have to skip the index read and use the first 6 bases of read 2 to get the barcode. MiSeq Reporter doesn't support this format for demultiplexing, so you'll have to manually demultiplex with third party software or using Bcl2Fastq with modified settings to generate the fastq files.

Alternatively, you can use a custom index read primer to make it compatible with MSR. The index primer sequence would be reverse complement of the sequence just before the barcode on the reverse primer: CCAGCAGCCGCGGTAAT. This will let you use MSR to demultiplex and generate the Fastq files. However, you'll still have a small problem of the 6 base barcode sequence getting stuck in the beginning of your read 2 sequence which you'll have to trim off.

Either way it'll take a bit of computational work to get the reads sorted and cleaned up. Whichever way you choose it up to you. The first method is probably easier to set up since you won't need to buy an extra oligo and fiddle with custom primers.