Issue with MiSeq Demultiplexing: Unexpected Ns in I5_Index

Question

[sergio.fern1994]

Hi BioStar community,

We have been routinely sequencing DNA samples using an Illumina MiSeq sequencer with CDI QIASeq Fx library kit. So far, we have not encountered any issues, as the MiSeq performed the demultiplexing process itself. However, in a recent run, the MiSeq failed to produce any fastq files except for the undetermined R1 and R2 reads.

To address this problem, we attempted to demultiplex the run ourselves using bcl2fastq2, but unfortunately, we encountered the same issue. Upon inspecting the Demultiplexing Stats, we observed that the I5_Index contains multiple Ns instead of the expected 8 bp sequence.

Most Popular Unknown Index Sequences
Columns: Index_Sequence Hit_Count
ATTACTCG+TNNNNNNN   59644
TCCGGAGA+TNNNNNNN   58629
CGCTCATT+TNNNNNNN   58571
ATTACTCG+ANNNNNNN   52028
TCCGGAGA+ANNNNNNN   51507
CGCTCATT+ANNNNNNN   49504
ATTACTCG+CNNNNNNN   44697
ATTACTCG+GNNNNNNN   41469
CGCTCATT+GNNNNNNN   41314
CGCTCATT+CNNNNNNN   40904
TCCGGAGA+CNNNNNNN   37623
TCCGGAGA+GNNNNNNN   36295

In this sequencing run, we used a combination of 701-703 for I7 and 501-506 indexes to assign 24 samples. Consequently, both indexes are required for successful demultiplexing. Because of this, we have come to the realization that this issue might not be recoverable from a bioinformatics point of view.

We are puzzled as to what could have caused this problem. Could you please share any insights or suggestions on potential causes for the unexpected Ns in the I5_Index during demultiplexing? The sequencer seems to have completed all cycles without issue, there were no errors or warnings reported and all 518 bcl files seem to be there.

You can find attached the flow cell & lane summaries:

You can also find the sample sheet used:

Thank you in advance for your assistance!

Answer 1

[GenoMax]

This run has a failed index read 2. There is not much you can do about this except check with Illumina tech support (you may have done this already). They can examine the run remotely or ask you to send in some log files. If this is a sequencer related issued (and if you have a maintenance contract) they will replace the sequencing kit and you will need to re-run.

Looking at the cluster density this run appears to be (borderline) overloaded and that may be the cause of the index 2 failure. Re-run at a lower loading conc.