bad quality of Read2 using a MiSeq ReagenKit v2

[mcnelson.phd]

What you're quoting for Read 1 sounds pretty normal to me, especially with amplicons. Read 2 will always be worse than Read 1, but the values you're giving are worse than what should be expected.

Have you looked to see if there is a marked difference in your metrics for the top and bottom flow-cell planes? If you're concern is that only one side of the flow-cell is being maintained/sequenced, then seeing if the metrics between the two planes would easily tell you if there really is an issue.

Another suggestion would be to try a phiX run to ensure that there really are no problems with your instrument.

As far as the low diversity goes, assuming you're system is updated you should be running the latest version of RTA, which deals with base skew pretty well. However, you are clustered above what Illumina would recommend, so try dropping your target value down to around 650 and try again. Lowering the cluster density to around that point should solve most low-diversity issues, assuming you're working with the latest RTA version.

[microgirl123]

Are you running the latest version of RTA on your MiSeq? That should deal with any problems arising from low diversity. And have you talked to Illumina tech support?

[MiSeq_Issues?]

Hello all,

is there someone else who uses the MiSeq Reagent Kit v2 with 2x251 cycles? I already started several runs for an amplicon-panel on the MiSeq and always the same problem occured: after 150 cycles in Read1 the Q<30 value rapidly decreases, however the average of Q<30 after Read1 is about 70-80% (Cluster Density (K/mm2) of 930+/-20), which seems to be okay. But the Q<30 for Read2 only starts at 50% and ends at about 30%. This results in bad sequences for Read2 with lower coverages and more dropouts and uncompleted regions. We already considered several issues that may cause this problem.

- The chemistry of the kit or a defective laser can't be the reason, because the Index-Read (containing 6 bases) which occurs directly after Read1 shows good results and a Q<30 of 98%.

- Another idea was that after a certain time our amplicons can't resist the mechanical pressure on the flow cell because of the wash steps etc. so that they break and consequently only areas situated on the flow cell basis could be read.

- Or is there something wrong with the instrument? a possibilitie in this case could be to ask Illumina for a prepared template. When all cycles (2x251) normally work, we can be sure, that the issue comes from our library preparation.

-Another point is the problem of low diversity that is common in amplicon sequencing: Actually our PhiX spike in was 25%, perhaps we have to increase the content to 50%... and to reduce the
cluster density by inserting less than 2nM (we actually add) from the library). We already designed modifications upstream of the primers to garantee more variation.

Perhaps this problem is not unknown and it happens to someone... I'm happy about each advice or idea,

thanks a lot in advance!