Which is more reliable - Tapestation or Qubit?

[ENT Fellow]

Hi there,

I'm planning to do some miRNAseq on human FFPE tissue. I've used the Qiagen FFPE miRNA kit to extract goal RNA including miRNA. On quantifying the RNA I get a big disparity between the Qubit concentration and the Agilent tape station concentration - why is this and which is more reliable. I used to do a Nanodrop concentration as well but found with previous DNA/RNA work that it always overestimated the concentration and the contamination ratios made no difference to the quality of sequencing library I made in the past so I don't bother with it anymore.

Any advice?

[NitaC]

I hate getting involved in what's "better" but in my lab, generally we do not pay attention to the concentration estimates from TapeStation. We use Nanodrop and/or Qubit. I have noticed though, that we use Qubit for concentration results that will ultimately be reported in a manuscript.
IMO, I prefer BioAnalyzer - I can get RIN's and concentrations that I trust all at once. The concentrations I get from BioAnalyzer are usually pretty close to what I get with Nanodrop.

But...I'm a bioinformatics analyst with some wet lab skills. So, please take my advice with a grain of salt.

[ENT Fellow]

Thanks for the response - I see your point of view - I have never used a bioanalyzer as just before I started we got the tape station and people loved it's simplicity and speed. It would be fascinating to see if anyone has compared bioanalyzer to tape station (concentrations, RIN etc). Amazing that you get similar bioanalyzer concentration to nano drop as we don't trust nano drop at all (usually overestimates concentration considerably in my hands) and use Qubit as gold standard (upto now). I guess the reality as as long as you get a library and it runs on the sequencer it probably doesn't matter.

[zrl]

Thanks for this question. I am dealing with the same issue. I have some total RNA isolated from a very small tissue sample. TapeStation claims that my samples are in the range of 800-1800pg/ul concentration and of high quality. However, when I tried to run Qubit (with the HS RNA kit) my samples are supposedly under the limit of detection, which is <250pg/ul. I am very surprised that both are so different. I would love to put my faith in TapeStation, but Qubit is preferred in my lab for quantification because it is supposedly more accurate. It's possible that I made a mistake with Qubit, but now I don't have enough sample to try it again. Any suggestions?

[zrl]

Qubit lower limits

I have figured out the issue that I was having with Qubit giving me an error message. Qubit has a lower limit of 20pg/ul final concentration for the RNA HS kit. (The lower limit stated in the manual is a higher concentration and probably reflects the lack of accuracy at concentrations below ~200pg/ul.) In order to fall above this cutoff, one needs to add at least 4ng of RNA to the 200ul of dye solution. Just remember that the lower limit applies to the diluted solution you put into the Qubit, not the starting concentration of your sample.

By constructing my own standard curve and using the raw values obtained on the Qubit calibration screen I was able to roughly calculate the concentration of my RNA. It matched fairly well with TapeStation, though some standards that I analyzed at 1, 2, 3, 4, and 5ng input RNA were variable in their accuracy. Hope this helps anyone facing this problem.

[mcampo86]

TapeStation Bioanalyzer
Length (bp) Concentration (pmol/L) Length (bp) Concentration (pmol/L)
Pool #1 435 354 405 950.1
Pool #2 438 376 418 575.1
Pool #3 461 437 426 972
Pool # 4 453 502 433 1080.6


The tapestation and bioanalyzer both give different fragment sizes and concentration values with the same set of samples. I have had issues with library pooling for PGM due to this difference. Which values can I trust? Both ladders were run, looked perfect. Everything was done as is of protocol. So whats the deal Agilent people??? Ive talked to other people with the same issue and no one from tech support knows why... and they know this is an issue with many customers

[kerplunk412]

The quality of data produced by the Tapestation is much worse than the Bioanalyzer, based on my own experience and conversations with others, including an Agilent rep. I would definitely trust the Qubit concentration over Tapestation.

Since you mentioned you are doing a miRNA-Seq experiment, I would also like to point out that there are big issues with ligase bias in most Small RNA-Seq kits. Bioo Scientific will be releasing a kit in June that reduces this bias through use of adapters with randomized ends. I don't want this to sound too much like a sales pitch, so please PM me for more information on the kit or the literature demonstrating ligase bias in small RNA library prep.

For full disclosure, I work at Bioo Scientific.

[nucacidhunter]

TapeStation Bioanalyzer
Length (bp) Concentration (pmol/L) Length (bp) Concentration (pmol/L)
Pool #1 435 354 405 950.1
Pool #2 438 376 418 575.1
Pool #3 461 437 426 972
Pool # 4 453 502 433 1080.6
The tapestation and bioanalyzer both give different fragment sizes and concentration values with the same set of samples.

The size variation seems to be consistent with CV documented in respective manuals. Quantification with both instruments is less consistent with PicoGreen for couple of reasons. Bioanalyzer uses 1 µl of sample and TapeStation 1 or 2 µl depending on Screen type. Variation in sample volume loaded and also buffer makes the quantification unreliable. In addition, except HS DNA 1000 Chip both instruments use the upper marker concentration to estimate sample concentration. If loaded sample contains fragments which are close to upper marker size, they will be mixed with upper marker and as a result the concentration will be estimated lower than real. In addition, CV of sizing and quantitation accuracy and reproducibility not only is different for various reagents but also varies in different size increments for the same reagent or kit. When I run amplicons less than 800 bp on TapeStation, I get good concordance between Tape and PicoGreen.

[ENT Fellow]

Hi there,

Sorry I didn't realise people had been responding. An update on my situation - I have tried making libraries using the same RNA sample using both the Qbit concentration and tape station concentration (in a sample which had a big disparity between the two results and in which the tape station reading was lower than the Qbit and have found that the Qbit gave a successful library, suggesting it is more accurate. This surprised me as my DNA tape station experience is that tape station and Qbit correlated closely and if they did not the tapestation was correct. Maybe it's just an RNA issue (mine is FFPE samples). I'm going to trust the Qbit for now. It may be an issue with the tapes - temperature, age, air bubbles all affect the tape station but it is still massively easier to use c.f. the bioanalyzer, so I am going to use the tape station to detect any adaptor contamination etc but going to trust the Qbit concentrations.One thing I noticed as well is the the tape station was almost unable to measure the concentrations in high concentration samples (~1ug/ul). It may be related to the markers in the tape station only being appropriate for a very narrow range of concentration to guarantee reliability. Sorry lots of speculation but the end result is for concentration I'm going with Qbit - and that's going in my thesis.

[ENT Fellow]

And a final update - I just ran 4 libraries on the high sensitivity DNA tape station. It read the highest concentration NA length as being ~190bp - a major problem. I ran the same libraries on the bioanalyzer which read the relative concentrations as the same (though absolute were lower on tape station) and the NA length as 142-152bp which is perfect. The tape station tapes were of course out of date as they go out of date so quickly everyone I know just continues to use them as per normal to save money but clearly they become inaccurate on judging the actual results. Just another piece of info which could have been obvious to many but not to me.

[northpaw]

You should check out the Fragment Analyzer from Advanced Analytical if you work with FFPE. They have a Large Fragment Kit that works really well for FFPE-DNA with a dynamic range from 1 bp through 20,000 bp and an upper marker designated at 100,000 bp. Its nice to have accurate sizing and quantitation in this large range because FFPE samples are SO variable.

Illumina uses the Fragment Analyzer (FA) now. They evaluated all systems on the market to improve reliability, throughput, and dynamic range of analysis for their workflows. This of course included the TapeStation. Illumina doesn't recommend the TapeStation, they have recommended the BA as you know for years.

Do a google search for the Illumina Automation Partners webpage. Scroll down and you will find the Advanced Analytical Fragment Analyzer (FA).

Illumina R&D has integrated the FA into multiple divisions including Consumables Product Development, Diagnostics, and Oncology. They are even starting to release new sample prep kits with the FA as a recommended QC platform.

The latest was the TruSeq RNA Access kit. They have a tech note for degraded RNA sample analysis that describes how they use custom software settings on the FA to streamline analysis of a new critical QC metric for RNA samples they call DV200.

You can find this in a google search on the Illumina website:

[PDF]Evaluating RNA Quality from FFPE Samples - Illumina
res.illumina.com/documents/.../technotes/technote-truseq-rna-access.pdf
Technical Note: RNA Sequencing. Introduction. The TruSeq® RNA Access Kit provides an exon-capture, RNA-Seq approach for difficult samples such as RNA …

If you work with FFPE or any samples prone to degradation by digestive enzymes, you should definitely check out this new kit from Illumina. Its pretty amazing.