General construction of double-indexed fusion primers for pair-end sequencing on MiSeq:
P5 (adapter) linker index i5 SBS3 (sequencing primer, read 1) template primer fwd AATGATACGGCGACCACCGA GATCT xxxxxxx ACACTCTTTCCCTACACGACGCTCTTCCGATCT ZZZZZZZZZZZZZZZZZZZZZZZ
P7 linker index i7 SBS12 (sequencing primer, read 2) templ primer reverse
CAAGCAGAAGACGGCATACGA GAT xxxxxxx GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCT YYYYYYYYYYYYYYYYYYYY
This construction is based on the TruSeq chemistry. At ICBR, they know it but it might be worth emphasizing when you submit your libraries. For more details on preparing a library using these fusion primers go here.
The steps of pair-end sequencing are as follows:
- SBS3 priming, read 1 sequencing
- i5 sequencing
- i7 sequencing
- SBS12 priming, read 2 sequencing
Therefore, the i7 index is the one that’s read first. We already have 96 versions of the read 2 primer, the one with the i7 index. The i5 index can multiply your multiplexing power.
Issues with base diversity
Illumina machines have trouble detecting signal from individual DNA clusters if all/most clusters have the same base across the lane. This is especially problematic with amplicons, where most sites will have the same letter in most reads. It can result in unclear signal. That is important to avoid especially in the indexes and at the beginning of the read when the instrument is still calibrating itself. There are two strategies to avoid low-quality reads due to homogeneity of DNA:
- Spiking with PhiX (a phage genome) – ICBR does that automatically, but you need to tell them that you have an amplicon, so they add more. That works, but it wastes sequencing power.
- Make the sequences mutually off-set, either by adding heterogeneity spacers, or by using sliding template primers. We use the latter. It is possible is the surrounding of the priming site is also fairly conserved. An example of a sliding LR0R primer (fungal LSU) is here:
priming site on template: 5'...GCCACCCGCTGAACTTAAGCATAT..3' original LR0R: ACCCGCTGAACTTAAGC LR0R(1): CCCGCTGAACTTAAGCA
The first PCR (to enrich for your fragment) should be done with the same uniform template primer, but these sliding primers can be incorporates into the fusion primers. They assure that the reads are off-set by one position which creates the desired base diversity at every sequencing step.
It is also very important to make sure that indexes re ballanced
Issues with read length
Here are ALL the barcodes/indexes you can choose from. Although the combination of two indexes should unambiguously identify the original sample, it is not a bad idea to use unique indexes throughout, both for the i5 and i7 segments: