MiSeq v3 low diversity

[thermophile]

I haven't tried a v3 2x300 kit in over a year. Back then it had worse quality than v2 2x250 at similar cycle (v3 cycle 200 was much worse than v2 cycle 200) and the extra 50 bases were very poor. Now I'm having issues with a particular set of 16s samples (other 16s libraries work fine, this one set of samples is failing in R2) so talking to tech support who is telling me that v3 is now optimized for low diversity.

Has anyone see an improvement in low diversity libraries on v3 kits over the past year?

[Tim NGS]

We noticed an increase of quality since march/april last year. On our MiSeq's we mainly run targeted resequencing (amplicon), so also (quite) low diversity. Even when using v3 2x300, almost all runs were above Illumina specs (mostly Q30 above 80%). Error rates also seem to have improved.

What I can suggest is to keep cluster density quite low, we target below 900-850 K/mm² with a PhiX spike in ranging from 3-5%.

In general quality of read 2, based on SAV, can still drop quite low a the end of the reads.

[thermophile]

v2 I generally aim to cluster 650-700 (though tech support is now telling me to go higher) and phiX and/or wgs 20% of the run

[Tim NGS]

I think you can aim for slightly higher cluster densities, but raising too much will definitely compromise on average quality.

Issues Illumina had with chemistry in 2015, are to our opinion, mostly solved.