Adding heterogeneity spacers to primers used for Illumina based amplicon studies has proven to improve read quality by increasing bp diversity on the flow cell during sequencing. I have seen two approaches to these heterogeneity spacers: add 0-(6-8) Ns or a sequence specific set of spacers. My intuition tells me that including a specific set of spacers is superior to just leading Ns - please correct me if I’m wrong. That being said, studies that include specific spacers do not or very minimally explain how or why they chose the specific sequences for the spacers they use. How does one design these spacers? I assume it has something to do with minimizing binding to the genome upstream of where the target primer binds, and avoiding sequences used in binding sites for Illumina primers.
Another common strategy is to just spike in about 20% PhiX control DNA.
But yes, you're right: don't use randomized bases, choose defined sequence for flanking. When I've done this previously, I basically randomly generated sequence but made sure that there was no significant alignment to any Illumina adapter sequences (including the indexes you used) to ensure that the sequencing primer had no problem with off target annealing. Also make sure that the sequence is 50% GC so that polymerases don't get stalled in your stagger sequence.
Another thing to check for is that you don't have any large homopolymerics stretches near the start of your amplicon. If you do, this can still mess with the optics of the sequencer even if you've added homogeniety spacers upstream. If you do have these present in your expected amplicon, you should just use the PhiX spike-in strategy, create extra long spacer sequences, or redesign your primer binding strategy.
Heterogeneity spacers are added to prevent the signal loss from the clusters as camera can get blind if all the clusters shine with same colour. Specific set of heterogeneity spacers as you say or random sequence is not a major factor here. Heterogeneity spacers of same size and same sequence will not be helpful in any case. If these spacers are different sequences (which is random any way; either by Ns or by random designing) or even of different lengths, they will be better to serve their purpose.
Theoretically, these spacers should not interfere the amplicon generation or sequencing, however, in case of specific markers, there might be some issues.
Abhijeet Singh the defined sequence works if the spacers are a mixture of different lengths. In other words, one should design 9 different primers for each direction that contain the first 0-9 bases of the known spacer sequence inserted between the gene-specific annealing region and the Illumina priming site and then combining them at an equimolar ratio. Doing this for the forward and reverse primers will create a pool of amplicons that will produce reads that all start at a different position in the spacer region.This will protect against focusing issues in the sequencer's optics. Sorry, this is something I should have made clear in my comment above.
When PCR is involved, I'm an advocate of keeping sequence as defined as possible where you can. Introducing randomized bases often leads to artifacts and biases.
- Badges
- Science method