Low intensity & quality for Index Read 2

[kmcarr]

We've encountered a problem with the last couple of paired end, dual indexed runs on our HiSeq 2500 and I wanted to check if any others have had similar experience. The problem is that the signal intensity and thus the read quality of Index read #2 is very poor. All other aspects of the run, meaning Read 1, Index 1 and Read 2 work perfectly fine, good signal and quality. The poor quality of Index #2 results in problems with demultiplexing.

We have run a couple of single read, dual index runs and in those cases index read 2 is fine, of course the manner in which index read 2 is done on single read vs paired end read flow cells differs.

We contacted Illumina and they have no explanation. Nothing stood out in the log files to suggest a cause. The engineer did some back flushing and checking of the fluidics but we have not tested it since that was done.

Does this sound familiar to anyone? Any ideas would be appreciated as well.

[LaoR]

Can you look the %base? See if you can read off the base pairs from it.

I've encountered something similar for 3 rapid runs.If you can read off the base pairs for your second index, search to see if it's from part of the I5 adapter sequence. If so, it may be a "transient" fluidics problem. The machine probably didn't deliver reagents to strip off the previous index sequencing primer.

[HeinKey]

We have the same issue. Low quality index2 in paired-end runs, and normal quality index2 in Single read runs.
Illumina finally recognized the problem, (we did some validation runs for them).
We now have a work around where we receive a different dual index primer and run a custom receipe on our HiSeq2500 (V4). This way the index 2 primer is not the grafted oligo, but the separately delivered index 2 primer which probably is covering the dark cycle nucleotides. We did not get the details of this primer sequence....
We will test this in the upcoming weeks.

[kmcarr]

We have the same issue. Low quality index2 in paired-end runs, and normal quality index2 in Single read runs.
Illumina finally recognized the problem, (we did some validation runs for them).
We now have a work around where we receive a different dual index primer and run a custom receipe on our HiSeq2500 (V4). This way the index 2 primer is not the grafted oligo, but the separately delivered index 2 primer which probably is covering the dark cycle nucleotides. We did not get the details of this primer sequence....
We will test this in the upcoming weeks.

HeinKey,

Yes, it sounds exactly the same as the issues we have had. At least now I know it isn't something totally unique to our instrument. The problem was so specific to paired end, dual indexed runs that I assumed it had to have something to do with priming index 2 from the grafted oligo.

We are starting another PE, dual index run today and we'll see how it goes. I'll report back and please do keep me posted on your results (you can PM me if your wish).

Thanks, Kevin

[LaoR]

If that's the case, wouldn't it be reproducible by using reagents from the same lot to run the libraries again using similar settings?

In the beginning, Illumina wants to blame on the library prep. We used same lot numbers to rerun the "failed index 2" run on the instrument. Rerun was totally fine.

[Ryjinko]

Our sequencer is NextSeq 500, and we encountered the same problem in paired-end dual-index sequencing of library constructed from PCR amplicon.

And now we have solved this question. During the library construction procedure, we use cutomized P5 and P5 adapters synthesized by commercial provider, by sequence analyis, we found that the adapter missed one "A" base when we designed, while this "A" base was added by dA-tailing in the standard Illumina library construction peocudure.

This is just the case for our NextSeq 500 sequencer using customized P5 and P7 adapters during paired-end dual-index sequencing.

Hope that will be of little help.