Illumina MiSeq Poor Q30 scores in Read 1 relative to Read 2

[jreuther]

Hi everyone,
We are consistently seeing an issue on our MiSeq where Read 1 has very poor Q30 scores, followed by a successful sequencing of Read 2 with Q30 scores above 80%. We have had successful runs where we didn't see this trend but it seems about every other run we see this happening. We have talked with Illumina Tech Support and they don't seem to know what is going on. They adjusted the temperature of the instrument (set it a few degrees lower) but this didn't seem to fix the issue. Has anyone else had similar issues? Any insight would be much appreciated.
Thanks! See below for run specifics:


Chemistry: Miseq v3 600 cycle kit.
Read configurations: R1 = 150 cycles, Index = 16 cycles, R2 = 150 cycles.

The library is a DNA hybridization capture, with avg library length ~320 bases

The library was run with 10% PhiX spike-in.

Custom primers were NOT used.

[luc]

Indeed very weird. Fortunately we have not seen this problem so far.
Are you using custom sequencing primers?

[nucacidhunter]

Two possible causes:
1- Instrument related
2- Library P5 adapter

In case2 the adapter sequences could have variations from standard sequences which affect R1 primer annealing or binding stability resulting in low quality. Paying attention to phasing/prephasing in good and bad runs should highlight some differences. This case is more prevalent in PCR amplicon based library preps that adapters are added by PCR because of the possibility that some oligos will have indels or base substitutions resulting from low quality synthesis or storage.

[jlli2000]

Does anyone have experience to run Miseq V3 600 kit with uneven r1 and r2 cycles? i.e. r1 400 and r2 200?

Thanks,

Jin

[Markiyan]

600 cycle miseq asymetric runs & seq alternatives for 600bp+ amplicons.

I had tried R1:319 I1:6 R2:300 and I must admit that as you approach 300 mark the quality decay (error rates increase) is quite exponential in it's nature :-(, so the reported quality at R1@319bp was lower than R2@300bp...

So it would be better to do R1:310 I1:6 R2:309

This is why the 2x400bp (800 cycle kit was never released :-(...
Also sometimes pooled v2 2x250bp kits (to make them 2x300) give better results than v3 2x300bp kits.

If you wish you can have your multiplexing index as I2, and do a custom primer read as I1 (if you want to sequence an area inside the amplicon next to the very conserved region):
Do something like: R1:210; I1:207; I2:8; R2:200 (adjust length as needed) or fragment amplicon by the nextera (long amplicon sequencing protocol, if your experiment can cope with amplicon assembly step)

If using a custom index1 - use a dummy I1 index sequence to specify the number of sequencing cycles for custom primer for "index 1".

PS:
Oherwise with 600bp+ amplicon (if you can stand 2-5% errors) do a nanopore 2D run.
(also may try a rolling circle replication from the amplicon insert with 2 hairpin adapters at the ends to generate a single stranded library for a multipass nanopore, but it would have higher systematic error rate than pacbio).

Or PacBio CCS (~<0.1% error rate with 4-6 passes but way less reads than nanopore).