MiSeq fungal contamination

[DRYTCYV]

Hello,
We detected some kind of biofilm on the Miseq manual sipper and have reported to Illumina support. The reply from the supposed Illumina support was a little bit unsupported. They said that they don’t have any official recommendation but we can use a bleach dilution to soak it in (manual sipper). They informed that tween solution can lead to a fungal growth, and we need to be aware that now all the lines (all sippers) can be contaminated, and there’s no official procedure for this kind of problems.
After reported several times some white spots in the images, looks like we discovered that it was fungus.
Anyone with the same problem?
Cheers,

[thermophile]

We had a film on the syringe plunger (didn't attempt to id it, it's either biofilm or chemical precipitation). I'm leaning towards precipitation because we do a bleach wash between each run. Illumina replaced the syringe and we stopped having runs fail due to bubbles.

[nangel]

Hi, we had noticed a similar fungal like growth on the side of the bottle containing the sipper after we had the machine in stand by for a few weeks. Although, I swapped all the solutions over before starting the sequencing run and had also done a bleach line template wash and 2 maintenance washes the next sequencing runs were very low in quality even with cluster densitys of around 800k/mm2 PF 94%. Illumina tech support had a look at the SAV files and have advised that fungal contamination is the issue here. Specifically the prephasing, high error rates, and the pattern of the Q scores dropping suggest this. Is this the same kind of run metrics you were seeing?

[DRYTCYV]

Hi nangel,

Thank you for your reply

Illumina tech support had a look at the SAV files and have advised that fungal contamination is the issue here. Specifically the prephasing, high error rates, and the pattern of the Q scores dropping suggest this. Is this the same kind of run metrics you were seeing?

Now there´s some metrics that make sense, the prephasing increase, Q scores problems etc. I'm gona check the last runs to confirm this.

Now the techsupport said to wash everything (all lines) with a bleach solution, but there´s no official protocol.

Cheers,

[JBKri]

Hi, we had noticed a similar fungal like growth on the side of the bottle containing the sipper after we had the machine in stand by for a few weeks. Although, I swapped all the solutions over before starting the sequencing run and had also done a bleach line template wash and 2 maintenance washes the next sequencing runs were very low in quality even with cluster densitys of around 800k/mm2 PF 94%. Illumina tech support had a look at the SAV files and have advised that fungal contamination is the issue here. Specifically the prephasing, high error rates, and the pattern of the Q scores dropping suggest this. Is this the same kind of run metrics you were seeing?

We too had some fungal growth in the wash bottle. We believe it may have been related to the water we used; the supposedly nanopure water turned out to have quite a lot of dirt inside. We switched to HPLC-grade bottled water; since then we haven't seen any fungus.

[nangel]

We are also doing the full line wash with bleach. My concern was that on going back to review all of the amplicon runs on the machine the pre-phasing and error rate were in exactly the same range from the very 2nd run on a brand new machine to the latest runs that were suspected to suffer from fungal contamination. I guess all I see is a rapid fall in the sequence quality as anything I can note as a difference. Still fingers crossed that fixes the problem. I guess the silver lining is that I now understand all of the run metrics much better and it should improve our run monitoring and loading (trying to keep that glass half full). I will also look at our water quality but we have changed SOP to autoclaving the MilliQ water in bottles and adding the Tween directly to this and keeping any stocks in foil in the fridge. With more rapid and recorded turnover, we are hoping this may also keep things ultra clean!

[ajeffs]

We don't do the bleach wash (yet), but did switch to keeping the MiSeq wash buffers in the fridge a few months ago after noticing stuff growing in the buffer bottles.

[docphil]

The quality of our MiSEQs runs has been consistently decreasing over the last 6 months, especially after the flip for paired end sequencing (read2). I am wondering what specific run metrics come with this fungal contamination and what kind of protocol are you using for a bleach wash for every line. Thanks!

[nangel]

We didn't see a huge improvement after the bleach washes or using another machine that should not have been affected by fungal contamination. I think our issue is something different after all. As far as bleach wash for the machine, I would make sure that the protocol came from Illumina tech support so that you are covered if it caused any problems with your machine. In essence it is just using diluted bleach in every position in the wash tray for 3 washes and then at least 10 MilliQ washes immediatley after. It was a very long day of washing!

Have your runs slowly been decreasing in quality especially for R2? Ours was a very clear change from previous runs to all the runs going forward after a specific date. There are so many run metrics you can look at but perhaps the clearest was looking at the %perfect reads in R2.

I should also say that we are only currently running 2 x 300bp V3 kits so we are never looking at the most fantastic R2 sequencing.

[docphil]

Yes we have seen a dramatic decrease in R2 quality, about a 30-35% decrease in %>Q30 compared to runs from ~6months ago. We routinely use V2 500cycle kits for 2x251 reads. My PI has spoken to a large commercial sequencing company who also use MiSEQ for amplicon sequencing and they have noticed a sharp decline in run quality in recent months as well. Perhaps illumina has changed the kit chemistry?

[DRYTCYV]

Hi everyone, sorry for the late reply and I need more time to check the previous Runs metrics.


I contacted tech support and they confirmed that the fungal contamination could be causing the white spots, that we saw in the images, and could potentially affect the metrics.

We started a new bleach protocol for all lines (1:100 dilution) and after that 2 water washes.

The last Run, was the first after the new wash protocol and looks like one of the worst runs in terms of quality. R2 Q30 ~57%. The cluster density was a bit high 1483 (V3 Kit), but i think this is not enough for a huge decrease in quality.

As far as bleach wash for the machine, I would make sure that the protocol came from Illumina tech support so that you are covered if it caused any problems with your machine. In essence it is just using diluted bleach in every position in the wash tray for 3 washes and then at least 10 MilliQ washes immediatley after. It was a very long day of washing!

Maybe we need to increase the water washes...

Have your runs slowly been decreasing in quality especially for R2? Ours was a very clear change from previous runs to all the runs going forward after a specific date. There are so many run metrics you can look at but perhaps the clearest was looking at the %perfect reads in R2.

I should also say that we are only currently running 2 x 300bp V3 kits so we are never looking at the most fantastic R2 sequencing.

Yes we have seen a dramatic decrease in R2 quality, about a 30-35% decrease in %>Q30 compared to runs from ~6months ago. We routinely use V2 500cycle kits for 2x251 reads. My PI has spoken to a large commercial sequencing company who also use MiSEQ for amplicon sequencing and they have noticed a sharp decline in run quality in recent months as well. Perhaps illumina has changed the kit chemistry?

R2 seems a identified problem that Illumina dont like to talk. To avoid a reduced R2 quality we need to keep the clustering arround 1000-1200 (R2 Q30~75%).

Cheers,

[nangel]

We only run our amplicons at 800 or below clustering with at least 10% PhiX. The recent decline in R2 quality has been confirmed by our FAS as a problem world wide with some batches of 2 x 250 bp and 2 x 300 bp MiSeq reagents. They are currently investigating. FAS have provided details of support and work arounds until the issue is resolved. Maybe check with your FAS to see if your reagents are similarly affected?

[dgovil]

DRYTCYV,

I concur with your statements. Our lab has been experiencing deteriorating Run quality on 2x250 and 2x300 kits.

the libraries that ran perfectly on older kits are failing and there are unofficial conformation regarding reagent quality from our FAS bit nothing official yet.

We have resorted yo running critical samples on the HiSeq instead.

Would love to hear other experiences as well

[JBKri]

DRYTCYV,

I concur with your statements. Our lab has been experiencing deteriorating Run quality on 2x250 and 2x300 kits.

the libraries that ran perfectly on older kits are failing and there are unofficial conformation regarding reagent quality from our FAS bit nothing official yet.

We have resorted yo running critical samples on the HiSeq instead.

Would love to hear other experiences as well

We also seem to have a problem. We just finished a 2x300 run, and we got only 68 % > Q30. The signal intensities were lower than our previous v3 run. The cluster density also turned out lower than expected from the loading concentration.

[Vinz]

Illumina currently has serious problems with MiSeq V2 and V3 reagents... Some lots perform terribly bad. Others are fine. It seems they are not communicating that well. I guess this is a problem particularly for smaller labs with lower throughput. Maybe it would be useful if we would come up with a shared spreadsheet where bad lots are identified?