MiSeq v3 600

[ATGCT]

Hi everyone,

we have troubles with our MiSeq Runs. At about cycle 80 we have a massive decrease in Q30scores.In general the quality of the runs is not good. Q30 scores of 50 or lower. In addition, during my last two runs- the intensity does not increase . I am very new to this technique and have therefore attached some Plots from basespace.
Maybe someone had the same problem? Or has an idea about what is going wrong- what are the possible causes? I went through the troubleshooting guides, but could find yet any answer.

thanks :-)

[GenoMax]

What kind of libraries are these? Are you spiking in phiX and if so at what %? Do you expect to have low nucleotide diversity after cycle 80 in sequencing? This run is also borderline overclustered and that could be one the reasons for this.

[HESmith]

The run is significantly over-clustered, as indicated by the low PF (only 68%; should be mid-80s to mid-90s) in combination with the high error rate (4.4%; should be <1%). Quantify your libraries by qPCR and load at the recommended concentration (actually, a bit lower for PE-300bp runs).

[thermophile]

Are you sure your template is more than 80bp?

[MU Core]

Is there a possibility of adapter dimers? May need to perform an additional library clean-up.

[HESmith]

Short templates/adapter dimers produce a drop-off in signal intensity, which @ATGCT doesn't observe (first thumbnail). And adapter dimers produce a diagnostic spiky signal, also not observed.

[nucacidhunter]

I wonder if you can post %Base from Data by cycle pan.

[ATGCT]

Hi, thanks for your answers !!!
We are doing Rad-seq libraries. Yes, I am quite sure, that my templates are bigger than 80bp- but we are checking this now using the bioanalyzer kit.
We already contacted Illumina, and they said, that we should increase PhiX concentration and lower our library concetration (from 8pM)- I think because of low diversity. I attached now % Base- at first it looks like problem of low intensities- then I dont know what happens

[ATGCT]

Ah yes @ GenoMax - we spike with 5% PhiX. I thin we should increase to 10%

[nucacidhunter]

Two issues:
1- low sequence diversity at the start which has resulted in low PF%. This should not happen if following standard RAD-seq protocol

2- short inserts and adapter-dimers that start reading into adapter in cycles 80-120

Both can be fixed by more careful library prep.

[GenoMax]

@ATGCT: You should take this data and use BBMap to estimate the insert size. You can do it in different ways. See this link for examples. You likely have very short inserts as @nucacidhunter has said. This will just confirm that observation.

[ATGCT]

Hi, thanks :-) I will have a look at it. But the insert size should be around 500 bp - I cut them out of a Agarose gel - so I have seen the size- unless something happens after gel extraction. We will have our Bioanalyzer results soon. Then I can also have a look at that. But I will try also BBMap.
Thanks for your suggestions :-) !!!

[GenoMax]

Don't be surprised. Keep in mind that small inserts preferentially cluster so if they are in there they are attaching to flowcell first.

[thermophile]

your %base also suggest short fragments and reading into the flowcell.