qPCR vs Qubit, big difference

[JBKri]

We have prepared a TruSeq DNA PCR-free library, and quantification with Qubit gives roughly 10 times higher concentrations than qPCR (range 4-50 times difference). Could this be because of poor adapter ligation efficiency, and is this normal for a PCR-free library? I was told by the local Illumina rep that such a big difference is not normal. We are using the KAPA Universal qPCR kit.
Jon

[pmiguel]

We have prepared a TruSeq DNA PCR-free library, and quantification with Qubit gives roughly 10 times higher concentrations than qPCR (range 4-50 times difference). Could this be because of poor adapter ligation efficiency, and is this normal for a PCR-free library? I was told by the local Illumina rep that such a big difference is not normal. We are using the KAPA Universal qPCR kit.
Jon

Sure, it could be low ligation efficiency. Could be an issue with your input DNA -- like maybe it is damaged. Maybe has a lot of abasic sites or whatnot that prevent many of the amplicons from amplifying. If you used a PCR-based method to construct the libraries you might not have even noticed.

Four-fold lower concentration via qPCR would be no big deal. But 50-fold?

Anyway, as long as you can get enough to cluster, it probably doesn't matter.

--
Phillip

[JBKri]

Sure, it could be low ligation efficiency. Could be an issue with your input DNA -- like maybe it is damaged. Maybe has a lot of abasic sites or whatnot that prevent many of the amplicons from amplifying. If you used a PCR-based method to construct the libraries you might not have even noticed.

Four-fold lower concentration via qPCR would be no big deal. But 50-fold?

Anyway, as long as you can get enough to cluster, it probably doesn't matter.

--
Phillip

thanks, these samples were prepared PCR-free. We believe the DNA quality was good. I fear we might have a problem with the qPCR workflow and/or instrument. We typically see quite a lot of variation between replicates in a qPCR. We will repeat the qPCR for this library.

[JBKri]

thanks, these samples were prepared PCR-free. We believe the DNA quality was good. I fear we might have a problem with the qPCR workflow and/or instrument. We typically see quite a lot of variation between replicates in a qPCR. We will repeat the qPCR for this library.

Just an update: we repeated the qPCR; it gave roughly the same results. We ran the library on our MiSeq; we got a low-ish cluster density (525 K/mm2) from loading 6 pM, so it seems the difference between the Qubit and qPCR is real and probably due to low ligation efficiency. The sample that had the extreme 50-fold difference actually had a mistake during the first bead cleanup (wrong ratio)...
Jon

[kerplunk412]

So then were most of your samples about 4x different between Qubit and qPCR? This seems more reasonable, but still fewer doubly-ligated molecules than I would expect.

[JBKri]

So then were most of your samples about 4x different between Qubit and qPCR? This seems more reasonable, but still fewer doubly-ligated molecules than I would expect.

The average was around 6-7x difference. Is this much less than typical?
Jon

[pmiguel]

The average was around 6-7x difference. Is this much less than typical?
Jon

Keep in mind people rarely check this and so won't give you a useful answer in most cases.

For no-amp libraries direct measurements of the amplicons themselves are necessary -- just use qPCR to quantify them.

I mean it's fine to sit around wringing your hands because only about 10% of the fragments you introduced into the assay actually produced amplicons, but the vast preponderance of people who do this sort of work would be blissfully unaware of their inadequacy in this area. So why sweat it?

Keeping it in the back of your mind is probably the best idea. Look for ways to improve the "quality" of your input DNA and look for issues with your pipettors, etc. But if you can scape together enough library to cluster then declare victory and move on. (This is for no-amp libraries only -- once you have PCR amplification involved in your prep methodology you should hold the final product to a higher standard.)

--
Phillip

[JBKri]

Keep in mind people rarely check this and so won't give you a useful answer in most cases.

For no-amp libraries direct measurements of the amplicons themselves are necessary -- just use qPCR to quantify them.

I mean it's fine to sit around wringing your hands because only about 10% of the fragments you introduced into the assay actually produced amplicons, but the vast preponderance of people who do this sort of work would be blissfully unaware of their inadequacy in this area. So why sweat it?

Keeping it in the back of your mind is probably the best idea. Look for ways to improve the "quality" of your input DNA and look for issues with your pipettors, etc. But if you can scape together enough library to cluster then declare victory and move on. (This is for no-amp libraries only -- once you have PCR amplification involved in your prep methodology you should hold the final product to a higher standard.)

--
Phillip

Yeah, that's pretty much how we feel, decent results for a first try, everyone is happy.
Thanks,
Jon

[kerplunk412]

The average was around 6-7x difference. Is this much less than typical?
Jon

Based on data I have seen from our kits I think only 1/6 or 1/7 efficiency is a bit low. However I agree with pmiguel that if you have enough for cluster generation you are probably good to go, especially with PCR-free libraries. I don't think adapter ligation to dsDNA has been shown to introduce much bias, otherwise I may worry about the low ligation efficiency.