MiSeq: no data for A/C but G/T look OK?

[sdbrown]

hi all,

i'm wondering if anyone else has seen this kind of data in their thumbnails before and if so what the problem turned out to be. it was sequenced with a v3 150-cycle kit; the first MiSeq we tried has a chronic pumping problem that illumina can't seem to repair and i ultimately drove the flow cell and reagent box to another instrument nearby that wasn't already doing a run.

thanks,
steve

[dannyhi321]

hi all,

i'm wondering if anyone else has seen this kind of data in their thumbnails before and if so what the problem turned out to be. it was sequenced with a v3 150-cycle kit; the first MiSeq we tried has a chronic pumping problem that illumina can't seem to repair and i ultimately drove the flow cell and reagent box to another instrument nearby that wasn't already doing a run.

thanks,
steve

Looks like a focusing/imaging problem - did you get any usable data?

[sdbrown]

Looks like a focusing/imaging problem - did you get any usable data?

hi, thanks for replying.

no usable data. there were a couple of tiles where A & C images looked OK too but of the ~9.9M reads reported by the SAV software there were zero reads passing the filter. i converted the .bcl files to .fastq files with bcl2fastq's --with-failed-reads option but still nothing.

edit: sorry, i should have said that the --with-failed-reads option produced FASTQ files with lots of reads composed entirely of 'N'.

[GenoMax]

In cases like this the first thing you should do is contact Illumina tech support. Have you done that? If they determine this to be a hardware/reagent failure then you could get a free replacement kit.

That said I don't remember seeing anything similar to images you have posted above. I wonder if one of the lasers failed on this sequencer.

[sdbrown]

In cases like this the first thing you should do is contact Illumina tech support. Have you done that? If they determine this to be a hardware/reagent failure then you could get a free replacement kit.

That said I don't remember seeing anything similar to images you have posted above. I wonder if one of the lasers failed on this sequencer.

that's exactly what illumina tech support said; A & C apparently share a (red?) excitation laser. i'm currently going through the back-and-forth with them but since the instrument is not connected to the internet, it's not as quick as it might otherwise be.

[GenoMax]

That makes sense. As long as the instrument is under a maintenance contract an engineer visit should fix that problem and you will get the free replacement kit.

[sdbrown]

is there a website somewhere that has an archive of instrument problems and the (diagnostic) data associated with those problems?

edit: in hindsight, perhaps seqanswers is that website...

[Jessica_L]

Illumina has previously hosted user group meetings with sessions for run monitoring and troubleshooting that discussed some of the diagnostic information you'd be interested in for run/hardware failures, but I don't know that there's an archive of any of that.

One thing I've noticed is that the grainy lines/static you see in the A/C images are some kind of an artifact exclusive to thumbnail images-- on older instruments (like GAs) where you could manually control lasers and nucleotide filters, if you went in and took some focus shots, you'd see a completely blank image, but when looking at the jpeg thumbnails they'd have a very static-y kind of look to them. I had asked Illumina at one point if it was related to the jpeg compression of the original tiff image file, but no one seemed to have an answer for that.

[GenoMax]

@Jessica_L: If I remember it right these images are saved with a non-standard bit depth and require a capable viewing program (e.g. Irfanview) that can decode them correctly.

[Jessica_L]

GenoMax-- the bit depth would make sense. But the jpegs from blank images have always looked strange, whether they were viewed in SAV or IrfanView-- I actually still have IrfanView on my computer, but maybe there's a setting I'm missing in order for the image to be displayed accurately.

[mgangel]

Dead red module

Just had the same problem. After a month of troubleshooting with tech support, we found that the red LED module was dead. I had runs fail at cycles 1, 16, 32, 176, and then I finally saw it die completely on the last run. They first blamed them software and downgraded to the previous version. They then replaced the hard drive. Finally they replaced the LED module for the A/C channel.