[subxero]

Thanks. Looks like I should use the HS kit since my DNA is only about 5 ng/ul

it's pointed out in another thread somewhere but if you are switching between the 2 kits you need to do some comparison tests of your samples using each kit. BR tends to give a higher reading by a high percentage vs the HS in my experience.

I once ran out of the BR dsDNA assay which I had been using to measure the concentration of my RNAseq libraries for cluster concentration calculations and instead used a HS assay in a pinch, did the calculations the same and clustered at "the same" concentration and all my libraries over clustered. I then did a side by side of the libraries with the BR and HS and there was on average a 29% difference between the values given, with the BR being higher. All the libraries were falling in the 40ng/ul range on the HS so they were well with in the BR kit range

I had good consistent clustering results with the BR assay when using it, and only had a problem with it when i switched to HS without making adjustments for the different values given. (<-- unknown to me at the time, or an oversight) I will stick with HS from now on for anything under 100ng/ul, I do feel like it is more accurate with in that range vs the BR

[lailaizhang]

Below 100ng then HS, 100ng-1ug then BR.
It's also good practice to re-QC normalised samples too.

Thanks. Looks like I should use the HS kit since my DNA is only about 5 ng/ul

[TonyBrooks]

Below 100ng then HS, 100ng-1ug then BR.
It's also good practice to re-QC normalised samples too.

[lailaizhang]

BR has been consistent for us. I only use HS rarely.

What is the range of DNA concentrations that you quantify? I am using Qubit BR assay kit to quantify genomic DNA of 5 ng/ul and I was wondering if I need to switch to the HS kit.

[Mr.Koyangi]

Hi,
Sometimes I like to run an 0.8% agarose gel with controls of known concentration with my samples and estimate the concentration based on staining of the sample. It may not be as accurate as picogreen, but it does allow you to see what is going on in your sample such as quality, possible contaminants (protein/RNA) and degredation, because picogreen will bind degrading dsDNA. But I find that most of the time the gel does correspond to other methods of quantification.

[pmiguel]

Hi creeves,
Thanks for bringing this up. I actually had some guys in our lab complaining about the viscosity of a genomic DNA prep they were working with -- thinking something was terribly wrong with it. Turns out they had concentrated it to > 1ug/ul, so yes, that is just the way genomic DNA is. And you are completely correct, you are never going to get an accurate concentration of the whole prep by taking an aliquot of that with a pipette.

I spent some significant fraction of my scientific youth doing genomic DNA preps in a lab where everyone did genomic DNA preps. So, to me it was completely obvious. Whole volumes of genomic DNA lore pass by my eyes every time I think about it. I literally just spent the last five minutes day dreaming about various genomic DNA prep methods of various resultant purities. And the attendant issues and even (tangentially) the old assay for determining the length of DNA molecules by the viscosity of the solutions it forms.

So many possible misunderstandings. Surprising protocols work at all, really.
--
Phillip

[creeves]

Genomic DNA quantification

I think I know what is going on here and I am surprised nobody has ever mentioned it in this forum. I just discovered it after working with DNA for many, many years, probably because the Next Gen protocols require such small amounts of DNA.

Picogreen or Qubit methods are definitely best for getting accurate quantitation, because they only measure the dsDNA. However, people still get variable results. So where is this variability coming from? I believe it is from trying to pipette small volumes of HMW DNA in order to make the required dilutions. When microliter volumes of HMW DNA are pipetted, some of the DNA (or even all of it, believe it or not!) can be dragged back out of the pipette tip as the tip is removed from the source tube. You think you just transferred 100 ng, but it might actually be much less. Why does this happen? Simple. Good HMW DNA preps are a massive tangle of strands. The higher the concentration, the worse the tangle. To ensure accurate quantitation, you must work with DNA solutions of lower concentration (no more than 0.1 ng/ul) and avoid transferring volumes less than about 5 ul when making dilutions.

By the way protocols usually warn of shearing DNA if you pipette too vigorously, but I don't think this is much of a problem. The above issue is much more serious.

[Rocketknight]

@Rocketknight, which assay of Qubit seems to give more accurate results, is BR assay consistent for your samples?

BR has been consistent for us. I only use HS rarely.

[TonyBrooks]

We also had this problem. We quantify using Picogreen (Qubit) and check integrity on a 2% agarose gel. Nanodrop usually over estimates (sometimes by as much as 10X) but we use it to check on 260/280 and 260/230 ratios.
We still had lots of discrepancy between what we quantified the DNA at and what our service provider estimated (even using the same method of QC). I think it's just one of those things.
In the end, we usually send 3-5X the recommended quantity and hope for the best. We still get samples dropping out but it's not as bad as it was.

We intended to look into DNA storage options (such as PureLink) to improve shipping conditions but never got round to it. Has anyone had any experience with such kits? Do they affect QC or library prep in any way?

[Eric@Kapa]

Hi Roxanne,

We (Kapa Biosystems) have a qPCR-based method for quantifying hgDNA that also provides a quantitative measure of DNA quality across a broad range of samples. We presented this system in a poster at AGBT earlier this year. Unfortunately, we have not yet launched a commercially listed version of the kit, but we are supplying some early adopters with custom versions, and we have an on-going test program for the kit.

If you would like more information, or if you are interested in testing the system, please let me know.

Best regards,

Eric

[Smriti]

@Rocketknight, which assay of Qubit seems to give more accurate results, is BR assay consistent for your samples?

[pmiguel]

We use Solid Taqman assays for our Solid machines. They have such kits for Illumina also. If I remember correctly, you can buy them from Kappa and Agilent. They are expensive, but then again so isn't Bioanalyzer chips. The advantage is that you are actually measuring amplifiable template, not just every DNA fragement in your sample. That is what really counts with Next Gen seq, are how many of those will become template for cluster generation or ePCR. BioA should be good to bring your multiplexed samples to equal conc. or to ballpark your library, but we always do a final qPCR before preparing our ePCR for sequencing.

I may be mistaken, but I took original post to mean that the samples being sent out were genomic DNA. (Eg, library construction being done by the core not by Roxanne's lab.)

If Roxanne's lab is constructing the libraries themselves, then I agree with you. Except I don't know of commercial qPCR kits for Illumina library Taqman. We use the SYBR green one from Kapa.

--
Phillip

[pmiguel]

My lab is a biobank that sends samples out to external labs for next gen sequencing. However, we find it nearly impossible to get matching results from our internal quantification methods and the lab that we are sending to.

My lab uses picogreen quantification to quantify the input genomic DNA, we also have a Nanodrop and an Agilent Bioanalyzer. We have found that our 3 pieces of equipment often disagree with each other and of course our measurements disagree with external lab's measurements.

Hi Roxanne,
In your case I would recommend that after your genomic DNA prep, you treat the DNA with a potent RNAse, like "RiboCruncher" and then deplete low molecular weight solutes somehow. We use a few cycles of dilution/concentration on Millipore Microcons, but they will not remove any contaminating high molecular weight proteins. So it might be worth looking into low X Ampure. I have not tried it with genomic DNA, so it might be problematic.

As touched on above, what is happening is that the nanodrop is confounded by all manner of substances commonly found in genomic DNA. pico green is probably giving you the right answer. But if your core is using another method they may flag the discrepancy. Agilent chips are useless for assaying full length genomic DNA -- genomic DNA is too long for the molecules to migrate past the detector before the run is finished.

Typical genomic DNA preps are >90% RNA and while RNAse will degrade the RNA into short oligo and mononucleotides, that will not decrease the amount of UV absorbed by those short oligos and mononucleotides. Further, even very low concentrations of phenol absorbs very strongly at 260 nm. That is, 0.1% phenol will absorb at 260 nm as strongly as 100's of ng/ul of DNA. And there are other substances that absorb at 260 nm.

So you need a method to degrade all the RNA and then remove it all, along with any other UV absorbing substance that is not DNA.

--
Phillip

[JeremyDay]

qPCR

We use Solid Taqman assays for our Solid machines. They have such kits for Illumina also. If I remember correctly, you can buy them from Kappa and Agilent. They are expensive, but then again so isn't Bioanalyzer chips. The advantage is that you are actually measuring amplifiable template, not just every DNA fragement in your sample. That is what really counts with Next Gen seq, are how many of those will become template for cluster generation or ePCR. BioA should be good to bring your multiplexed samples to equal conc. or to ballpark your library, but we always do a final qPCR before preparing our ePCR for sequencing.

[arnoldH]

You should be fine using all these three methods. Did you ever try to compare the references your service provider is using for fluroesecent measurement against your PicoGreen measurement?