Low miseq clustering with custom primers

Hello all!

This is not directly a Mothur problem, but I’m posting here in the off chance someone else has run across this problem and might have some insight.

My lab uses the V4 custom amplicon and sequencing primers detailed in the MiSeq wetlab SOP on the Schloss github. We have used these primers for ~4yrs and have always had good results with them. Recently, however, we have been getting really low cluster densities on the MiSeq machine (we would spike in to another run with 30% of the DNA pool and would only get back 1% of the total reads). If we use an entire flow cell for our 16S data, we will get reads but the run is still markedly under-clustered. The reads we do get are of good quality, however. This started a year ago when our MiSeq stage temperature was re-calibrated (it was too low before so the technician reset it higher).

The Illumina reps believe this is due to the custom primers we use for 16S library prep and/or sequencing (despite having used these primers ~20 times in the last 4 years without a problem). We did Sanger sequence a library through the amplicon primer and do see a small region in the adapter sequence that Sanger always has trouble with (perhaps due to secondary structure). But, again, these primers have always worked for us in the past and we can’t change the adapter sequence regardless.

Has anyone else run into this issue with the Schloss custom primers?! I’m guessing at least some of the people on this forum use these same primers…Any help would be greatly appreciated.


We use the same primers in our lab without a problem and have been getting pretty regular servicing on our machine. Perhaps the primers have degraded? These are expensive experiments to troubleshoot, but could you redilute your stock primer solutions and try again?


Thanks for the response!

Unfortunately (or fortunately?) we’ve tried two different sets of primers, one old and one new, and both have failed. Both we re-diluted from stock. We also added a library that we know has worked in the past to the most recent run and it didn’t work as well…We don’t HPLC purify our primers (which the SOP says you don’t either?), but we might try for a few barcodes just to see if that helps.


Hmmm. We don’t HPLC ours either. Might be good to have the technician back out to double check the calibration?

yup have seen it. The annealing temp of the Kozich sequencing primers is lower than the cluster generation annealing temp on the MiSeq. What I did was order sequencing primers with locking nucleotides to raise the annealing temp without changing the bases. shoot me an email at mars at uconn.edu if you want to see where I put the locks, it probably doesn’t matter, you just need to get the annealing temp up to (I think) 65. The temp difference is particularly problematic for the index and read 2.

I found my order, here’s where I put my locks. I mix this with the non-LNA sequencing primers (so I spike in 3ul of each primer). I ordered this from exiqon in 2016 for ~$250. The quality of my R2 drastically improved with these primers.


Thank you @Kendra!! I’ll order and post again with results.