MinION Flow Cell Issue

[KKaplan]

Hello everyone! My name is Kramer Kaplan, and I am a current graduate student working on sequencing mitochondrial DNA for freshwater stingrays using a 1D protocol for the MinION. My advisor and I had a few questions/issues we wanted to present to the community. I prepared a pooled library for four individual fish (2 species each) with four separate barcodes and after following the protocol, I checked the final concentration of the library before it was loaded onto the flow cell, and it was 13.5ng/ul, well above the 4.0ng/ul requirement for a 1D protocol. When my advisor and I loaded it onto the flow cell, the QC looked fine, but the "Strand-to-Pore" ratio was around 5%, and it should be in the 85-90% for this procedure. We were wondering if this was a simple library quality issue, or was this a flow cell issue. The flow cell is older, it was ordered at least 6 months ago, and I know that they tend to work best within the first 3 months of being ordered. It appeared as that the DNA was not being pulled through the pores, so it is possible that the proteins binding to the DNA in the pores are just not working due to an older flow cell, or does that not depend on the flow cell's age? Or, is it a combination of flow cell and library quality? We plan to order a new flow cell in the meantime anyway. Any advice/input the community could provide would be greatly appreciated. Thank you all so much!

[pmiguel]

What assay did you use to determine the concentration of the library?

--
Phillip

[KKaplan]

We used a Thermofisher Nanodrop-Lite.

[tbanks]

Hi Kramer,

When you ran the flowcell QC how many active pores were detected? 5% 'in-strand' pores is very low and is likely a problem with your library. The system is very sensitive to contaminants in your DNA. If you're using the rapid 1D kit it's even more sensitive.

ONT recommends you quantify using a Qbit. Perhaps others can comment on the relationship between Qbit and Nanodrop concentrations. Oxford also has some recommendations on 260/280 and 260/230 ratios but in my hands those values can't predict if the DNA is going to do well. I think Phillip has a few posts on this subject.

I'd recommend trying the lambda control library prep so you can get the feel for what a good run looks like. You can wash the flowcell after it has run for a few minutes. If you are using the 1D ligation kit I'd suggest extending the end repair and adapter ligation incubation times to those recommended by NEB (ONT cuts those times significantly). If you are using the rapid kit I would consider running a prepared library on a low % agarose gel, you should see fragmentation around 4 kb if your library prep worked.

DNA quality is the key with this technology. Good luck!

Travis

[pmiguel]

We used a Thermofisher Nanodrop-Lite.

Yeah, you don't want to use a spectrophotometer to measure library concentrations. DNA isn't the only molecule that absorbs UV at 260nm. Unincorporated nucleotides and RNA will, of course, also absorb there. Most genomic DNA preps end up 30% RNA, at least, unless heroic measures are taken to remove the RNA. It is not uncommon to see genomic DNA preps that comprise less than 10% DNA based on A260 measurement.

Please note that RNAse, even if it degrades RNA in a sample down to mononucleotides will not substantially alter the amount of UV absorbed by the RNA in the sample.

Finally, if you used phenol to extract your genomic DNA, then you should be aware that even trace (eg, 0.1%) phenol remaining in your sample will absorb so strongly at 260nm that it will likely make it impossible to determine the actual concentration of DNA in your sample. (Actually the newest version of Nanodrops claim to be able to correct for phenol. So if you have one of those, then a little phenol may not throw the reading off.)

Anyway, the basic story here is that UV spectrometers are not a good way to determine the concentrations of genomic DNA samples, nor libraries that derive from those samples. Fluorimetry or even a gel with a mass standard will give you much more accurate results. I have some more details in this thread.

--
Phillip