MiSeq Run, Problem with Index reads

[spf2]

I just ran a miseq run using NEBNext Ultra DNA prep, (the bio analyzer results looked fine prior to sequencing).

We sequenced 6 samples ( 3 samples X 2 duplicates) all with different index primers (NEBNext primers for illumina 1,2,3,4,6 and 12#). However, of these six samples, three were substantially low in concentration below normalization. These three contained indices 4,6 and 12. Using the sample sheet from illumina, it said that the indices 1,2,3 would be invalid if you used them together, however for this reason we included the samples with indices 4,6 and 12 despite the low quantification. (we know we shouldve used a better combination but the miseq reagents were already thawed)

The miseq was run, and resulted in overclustering, no (0) index reads for the 6 indices, and no clustering for the second read. This resulted in single file of undetermined reads.

The questions i have are, does any one know what the most likely reason for the failure? the Overclustering? the lack of any index reads? And is there any way i can dive into the fastq file generated and see my DNA was even sequenced or if its just random stuff?

Thank you for any help!

[GenoMax]

Do you know what the cluster concentration was? Is this a version 2 or version 3 reagent run?

[ScottC]

Without seeing the rest of the run details, I'd guess that it was due to the over-clustering. I would expect to see a very rapid quality drop-off too, in read 1, despite an 'OK' start.

Bear in mind that the cluster number reported by MSR might suggest that it's only slightly over-clustered, but it could be much more severe than it looks. You can look at the cluster images for an idea of how heavily clustered the flowcell was. MSR can't count the clusters accurately when they're too crowded, so the reported number isn't an (always) an indication of true density... or library concentration. Don't use this number to dilute your sample if you re-run it.

[nucacidhunter]

Over clustering can affect correct reading and recognition of indices and lowers the number of reads that can be parsed. To my knowledge if the first index read fails and software is not able to find any index, MiSeq aborts sequencing and does not proceed for second read. Illumina Tech support can have a look at run files to see why the run has stopped and if it is related to lack of index recognition.

You can map (if there is a reference genome) or BLAST fastq file reads to see how many of them are related to your species

[pmiguel]

Over clustering can affect correct reading and recognition of indices and lowers the number of reads that can be parsed. To my knowledge if the first index read fails and software is not able to find any index, MiSeq aborts sequencing and does not proceed for second read. Illumina Tech support can have a look at run files to see why the run has stopped and if it is related to lack of index recognition.

I don't think a MiSeq will abandon a run because of poor index demultiplexing.
I've seen cases on both the HiSeq and MiSeq where extremely over-clustered flowcells produced a low quality first read, poor indexing and then no real results from the reverse read. I think this might be because at extremely high cluster densities the first read may use up all the flowcell oligos, not leaving enough around for turn around to occur.

Also, while indexes work a bit better if you have a good balance of bases at each position, they still work to some extent even if you have a terrible balance. So it is definitely the high cluster density causing your issues.

Of course, just because you have one problem, doesn't mean you won't have other problems as well. "When it rains, it pours", eh?

--
Phillip

[geneart]

Hi all, on the other end , using NEBNext high fidelity kit I had libraries made. Used 1% phiX. But miSeq failed giving an error saying "clutering failed". There is absolutely nothing to see why or where it failed? It had happened before but then the PhiX 100% run had worked, to check the machine and illumine told that there was nothing wrong with the machine. Is it possible there is some problem with the enzymes that there is no cluster being formed? Or could there be a problem with the indices? I am new to this and any suggestion might help. Thanks

[Jessica_L]

Hi all, on the other end , using NEBNext high fidelity kit I had libraries made. Used 1% phiX. But miSeq failed giving an error saying "clutering failed". There is absolutely nothing to see why or where it failed? It had happened before but then the PhiX 100% run had worked, to check the machine and illumine told that there was nothing wrong with the machine. Is it possible there is some problem with the enzymes that there is no cluster being formed? Or could there be a problem with the indices? I am new to this and any suggestion might help. Thanks

Cluster generation failures can result from several steps in the process, including carrying over too much NaOH from the denaturation step. Likewise, if you're not using a fresh NaOH solution (check the pH of it, always), your template may not have denatured sufficiently for flow cell hybridization.

You might try double checking your NaOH first. To help you further, I think you would need to post some of the QC data for your libraries-- i.e. concentration from qPCR and a copy of the bioanalyzer results.