MiSeq bright spots in "C" images only.

[pmiguel]

MiSeq run clustered at low density (~450 K/mm2).
After 8 cycles, panel one looks normal:



But at cycle 10 a bright spot appears in the "C" image only:

It persists through later cycles.


Then at cycle 59 an additional bright spot appears. Again only in the "C" images:

It also persists through later cycles. Including through the index and second read.

What are those? I see them in another panel (different position in the panel) as well. Same size.

--
Phillip

[aleferna]

Any progress on these I'm getting the same problem?

[aleferna]

Even more so I'm getting these weird effects

[ECO]

Aleferna: your problem looks dramatically different to me (on my phone ha!), more of a complete lack of clusters (or lack of signal in that channel due to a low complexity library or adapter dimer, etc). Do all 4 channels look the same?

[pmiguel]

No. Actually the run was overall pretty good, so I guess I forgot to report this issue to Illumina. Yours looks much worse. You definitely want to report it to Illumina tech support.

Let us know what the diagnosis is...

--
Phillip

[pmiguel]

Aleferna: your problem looks dramatically different to me (on my phone ha!), more of a complete lack of clusters (or lack of signal in that channel due to a low complexity library or adapter dimer, etc). Do all 4 channels look the same?

I just assumed the images Aleferna posted were much lower resolution. But maybe that is something completely different...

--
Phillip

[aleferna]

Yeap, I've been looking at the raw files and I see these "blobs" / "flares" all the time, where do they come from?

On the other hand, I think we could be having a problem with adapter dimers, we didn't have any problems with the original nextera, but these Nextera XT is giving us problems. I'm waiting for the response from techsupport but I wanted to get a second opinion in the mean while.

Why would primer dimers cause these diming problems, we're having problems getting rid of these? Basically we're having lots of problems with the quality of the second read, the first one is sort of ok (not as good as the PhiX though), it looks as if we are loosing signal intensity too fast, could primer dimers cause this?

Pardon my ignorance its just our second run... I had never seen these images before.

[pmiguel]

Not sure what causes the bright circles.

If you want our opinion on primer dimers, I think we need to see some bioanalyzer images.

About the losing signal intensity, that sounds like a fluidics issue--to which MiSeqs are prone. But the MiSeq does have a good diagnostic utility with tests of various sub-systems. If you give tech support a call they can walk you through it.

--
Phillip

[Richard Finney]

I'm curious.
Are the blobs detected and then 1) filtered out ... OR .. 2) tagged as bad quality ?

These images remind me of the blobs found in microarrays ...

http://bioinformatics.oxfordjournals...expansion.html


""
MICROARRAY Blobs (not nextgen)

[aleferna]

well even if they are flitered out, it will introduce a heck of a lot of N's in a large number of reads, not sure how big that pic is but the blob is on top of a lot of clusters. Why only C, well I do have these "flares" or "bubbles" (I hope they're not air bubbles) in a few cases in the A track but nothing in the G or T???

[mukeshwar]

Hello Everyone,

I am also facing the same problem specifically in channel "C". Did you guys get any answers from Tech Support.

[pmiguel]

I don't think I ever talked to tech support about these. If you do, please post what they say.
--
Phillip

[bunce]

Hi All,
I have seen this only once in 100 MiSeq runs - I was told... " xxxx asked me to get in touch with you regarding washing the MiSeq after bright spots have been seen. We don't have a documented workflow as it is just a maintenance wash performed with warm water only. This does a really good job of clearing the lines of any crystals etc.

Please let us know if you do continue to see bright spots on your next run, please bear in mind that an occasional spot isn't unusual and shouldn't have much impact on the data."

So I think this is some kind of blockage - we did a warm water wash (as suggested) and have not seen it since. Even the run with the 'hot spot' seemed fine.

Hope this helps?, Cheers Mike