[memento]

For the custom libraries we had a hard time using RT quantification (always overestimating). We were using PhiX as a standard though. Currently the best way for us is to use gel cuts followed by PicoGreen (dilute if needed). It works well if Your library has unified size distribution.

We also had problems with v1 valve (it never failed completely, it was sort of intermittent getting worse over time) . I noticed that after they replaced it (and all lines) our clustering became consistent.

[abyss]

Hi Memento,
I agree with your observations and have observed a similar trend.
I thus stick with the Qubit measurement, which I believe uses a Picogreen or similar kind of a dye, and is quite accurate.

How would one get to check if the valve is a problem? The machine is relatively new, and not too many runs have been conducted (probably 10 -11 runs).
Would I need to do a PhiX flow cell? Does Illumina help with providing a test kit?
Thanks

[abyss]

The solution to the problem comes definitely by controlling the NaOH concentration to a max of 1mM. I tried loading the libraries at 10pM concentration with 1mM NaOH final concentration and that did the trick.

[LabRat89]

Hi everyone,

So I understand that the final concentration of NaOH should be a max at 1mM. However, we've been loading it at 1.2mM final concentration with 12pM sample input. Our cluster densities have been roughly 500k/mm2 for the past 3 runs. Would you say the 0.2mM difference is negligible?

Thanks in advance!

[nucacidhunter]

Illumina's stand on max of 1mM NaOH is pretty firm, so it must make a difference. It is easy to achieve this and follow Illumina's instruction. If library concentration is low it can be concentrated. You can also increase loading to 15-16 pM (maybe in small steps). Just have to make sure that low cluster number is not due to over-clustering where RTA struggles to recognise very closely located clusters and do not use them for template generation.

[bilyl]

Illumina's stand on max of 1mM NaOH is pretty firm, so it must make a difference. It is easy to achieve this and follow Illumina's instruction. If library concentration is low it can be concentrated. You can also increase loading to 15-16 pM (maybe in small steps). Just have to make sure that low cluster number is not due to over-clustering where RTA struggles to recognise very closely located clusters and do not use them for template generation.

Has anyone ever tried using the NextSeq style of denaturing libraries for the HiSeq/MiSeq? You neutralize the NaOH with an equal amount of 200mM Tris HCl pH 7, then dilute with HT1.

[LabRat89]

Illumina's stand on max of 1mM NaOH is pretty firm, so it must make a difference. It is easy to achieve this and follow Illumina's instruction. If library concentration is low it can be concentrated. You can also increase loading to 15-16 pM (maybe in small steps). Just have to make sure that low cluster number is not due to over-clustering where RTA struggles to recognise very closely located clusters and do not use them for template generation.

Thank you for the response! I ran a different sample set and got roughly 733k/mm with 14 million reads. This is still really low on the V2 considering how we're supposed to get 18M reads. After upping the concentration to 16pM, the cluster density generated was even less at 599k/mm. I'm guessing there's over-clustering now?

Additionally, has anyone used kits past expiration date and still have it work? If so, what is the time frame?

Thanks again!!

[nucacidhunter]

Cluster densities given for Illumina systems are for sequencing libraries prepared with Illumina kits and does not apply to third party kits or methods. This attachment is a useful source for troubleshooting cluster densities issues:

[LabRat89]

"Needs Attention" Message on BaseSpace

Thank you nucacidhunter for the attachment.

I encountered another problem that I've never seen before. Under BaseSpace, the run is showing "Needs Attention" even though the MiSeq instrument displays successful completion of the run.

Please see attached image. Have you or anyone else encountered this before?

Thank you again!

[NextGenSeq]

The low cluster density is usually caused by your NaOH. Either your NaOH pH is off or you are adding too much Na+ ions. which interfere with library hybridization to the flowcell.

We always check the pH of our NaOH prior to using it by making up a 0.1 M solution of our stock. The pH of 0.1M NaOH should be at least 12.6 or higher.
CO2 from the atmosphere will acidify the NaOH solution over time. Also, make sure you are using the more recent denaturation protocol (attached).

Denature DNA for 4 nM Library
1 Combine the following volumes of sample DNA and freshly diluted 0.2 N NaOH in
a microcentrifuge tube:
• 4 nM sample DNA (5 μl)
• 0.2 N NaOH (5 μl)
2 Discard the remaining dilution of 0.2 N NaOH or set aside to prepare a PhiX control
within the next 12 hours.
3 Vortex briefly to mix the sample solution, and then centrifuge the sample solution to
280 × g for 1 minute.
Denature and Dilute DNA
Preparing Libraries for Sequencing on the MiSeq 9
4 Incubate for 5 minutes at room temperature to denature the DNA into single strands.
5 Add the following volume of pre-chilled HT1 to the tube containing denatured DNA:
• Denatured DNA (10 μl)
• Pre-chilled HT1 (990 μl)
The result is a 20 pM denatured library in 1 mM NaOH.
6 Place the denatured DNA on ice until you are ready to proceed to final dilution.

[pmiguel]

The low cluster density is usually caused by your NaOH. Either your NaOH pH is off or you are adding too much Na+ ions. which interfere with library hybridization to the flowcell.

We always check the pH of our NaOH prior to using it by making up a 0.1 M solution of our stock. The pH of 0.1M NaOH should be at least 12.6 or higher.
CO2 from the atmosphere will acidify the NaOH solution over time.

I don't think the Na+ ions cause the issue. It is the OH- ions that are problematic above a certain concentration.

--
Phillip

[PrestoMKBC]

Additionally, has anyone used kits past expiration date and still have it work? If so, what is the time frame?

LabRat89 - I didn't see a response to your last question, but thought I'd chime in - we have tested a couple of "expired" MiSeq kits for a validation experiment, and after a month, you start getting clustering errors. We had a support Tech looking at a different issue with our machine, and he was okay using a flow cell that was a month old, but not the cartridge.

[lpalacios]

Dear all,
I am having problems with cluster density in the last 8 libraries prepared with Agilent Sureselect XT kit. I have run more than 40 libraries prepared with this kits with an average of 800 clusters when I loaded 8pM but in the last runs they have decreased to about 250-300 clusters! I have tried reducing NaOH in the denaturalization, the protocol starting with 2 or 4nM, to load 8, 12.5, 15 or 20pM of library... And nothing worked. I have run other libraries prepared with other kits and they work ok, so it is not an issue of the MiSeq. The final purification is done with agentcourt beads and the bioanalyzer traces are ok, similar to runs that worked fine.
Could anybody help me?

Thanks a lot

[Jessica_L]

lpalacios-- if you're fairly certain it's the Sureselect kit, I'd consider the number of freeze/thaws the kit components have been through. If you're using a brand new kit, maybe check whether this is a new kit lot number compared to the other ones you've used. I would assume, though, that if the problem was with the kit, you'd see some problems on the bioanalyzer-- reduced yield, etc.

The other question to ask is how are you quantitating your libraries after construction? Are you using a qPCR method?

[lpalacios]

Dear Jessica, thanks for your replay. I am not absolutely sure about the problem is the agilent kit but other libraries prepared with other kits and quantified and denatured in the same way are working properly. We have prepared 2 batches of libraries with 2 different lots and both failed, they were not used before. In fact, we have repeated 4 samples with a new kit sent by techsupport and we have no clusters... The quantification has been done with Qubit and bioanalyzer, in the same way as the 200 libraries that worked fine in previous runs. In fact, we have tested several concentrations and all of them had the same low number of clusters. Perhaps there is somthing in the pcr2 primer sequences, something that makes the DNA more sensible to denaturation... I don't know but today I have launched a new run and I have obtained low clusters again...