Covaris vs. Biorupter

[ETHANol]

For the preparation of chromatin for ChIP-seq, what is the advantage of the Covaris over the Bioruptor that warrants the big difference in price of the two machines?

[Hamid]

Hi ETHANol,

The two instruments utilize very different technologies for processing samples. I will try my best to simply describe the technologies.
The Covaris AFA utilizes spherically designed 500KHz acoustic transducer which focuses the 1mm wavelength generated acoustic energy to a small focal point located within the sample tube. In comparison bath type sonicator such as the BioRuptor operate in the 15-20KHz range and generate wavelengths of 10-15cm long. The unfocused long wavelength acoustic energy is dispersed throughout the water bath and absorbed as heat energy. That explains the reason why ice must be fed to the water bath, and samples allowed cooling down time after short bursts of sonication. Consider the area and volume of the water bath sonicator as compared to the small volume and area of the tube containing your sample. Only a very small fraction of the acoustic energy generated is targeted at your sample.
The focused millimeter wavelength acoustics of the Covaris AFA is targeted within the sample contained in a closed tube, and it does not generate heat, and therefore provide an isothermal processing environment for your samples. It also does not require rest periods to cool the samples down. Thermal control is quite important for DNA/chromatin shearing since a lack of thermal control can lead to thermal degradation, and ripping-off of the epitope.
The short millimeter wavelength of the acoustics, along with advanced electronic, provide unprecedented control over the sample processing. This level of control and focus, has allowed Covaris to provide validated, highly reproducible, non-contact protocols for shearing DNA in the size range of 100bp to 5kb. We also have validated protocols for chromatin shearing using SDS based, and non-ionic detergent based buffers. The same level of reproducibility and size range control is available for chromatin shearing as well.
The Covaris distributor in Europe is KBiosciences (http://www.kbioscience.co.uk). They can provide you pricing, possible demo opportunity, and further information.
If I can provide you any further information, or a more detailed description of the Covaris AFA technology and related applications data to help make an informed decision, please let me know.

Thank you

[ETHANol]

Hamid, thanks for the detailed reply. It's clear that technology behind the Covaris instrument is more advanced. What is not clear to me is if it has any practical advantage over the Bioruptor.

What I would be really interested in is if one technology over the other produced a more random shearing of DNA X-linked DNA.

[Hamid]

Hi ETHANol,

Sorry for the delay in responding.
It has been shown that lack of thermal control does lead to biased DNA shearing in the AT rich areas which are prone to melting more easily as the processing temperature is increased. Single stranded DNA is shared more easily than double stranded DNA. That explains the obsession with keeping DNA samples cold during sonication but adding ice chips, having rest period between sonication bursts, and even shearing DNA in the cold room.
As far as cross linked DNA is concerned, the lack of thermal and acoustic control does lead to striping of epitopes from cross linked DNA. We recently were provided some data that shows the difference in TaqMan of 100kb region of Murine IgH locus of chromatin sheared using a sonicator and Covaris AFA. Not only was the enrichment 10-20 folds higher, they also found epitopes they had not seen before. I will post the data as soon as I have their full permission to do so.


Thank you

Hamid

[captainentropy]

any word Hamid?

We're having a raging argument here on the advantages of Covaris over Bioruptor. I'm being told things that frankly seem ridiculous but I don't want to be too hasty.

One thing I was told by my PI who talked to someone at Helicos is that the Bioruptor makes DNA ends largely unrepairable, which seems unlikely since the first step of the library protocol is to repair ends, and we do in fact get lots of reads from our libraries so that can't be completely, if at all, true. The person also said the Bioruptor can remove 5' phosphate groups, which, even if true, is also repaired by the PNK in the end-repair step.

[Hamid]

Hi Captainentropy,

I really like that screen name!
Helicos does use our technology for shearing. I am sure that they have conducted testing on different shearing methods, and perhaps base that statement on some in-house results that I am not aware of. I will check with them on that.

Meanwhile, is there any specific information that you require?
Incidentally, in a few days I will post a case study from a data set from the University of Toronto in which they prepared libraries with identical DNA sheared by Covaris and Bioruptor. Their analysis of read depth(frequency vs read depth)clearly indicated a bias in the Bioruptor sheared samples whereas the Covaris processed samples generated a near perfect Gaussian distribution.
Have you sequenced Bioruptor sheared DNA and compared to Covaris? If so, i would love to get some of your raw reads for analysis.

Thank you

Hamid

[Hamid]

Hi Captainentropy,

Here are some advantage of Covaris:
1. Isothermal process. No need for adding ice chips, or processing in a cold room.
2. Dynamic range of DNA concentration as starting material. There is no minimum DNA concentration limit.
3. Regardless of DNA concentration, the same settings will generate the same fragment size range. No need for running a time course to optimize shearing.
4. very fast total processing time of DNA shearing. 200bp in 3 minutes, and 300bp in 80 seconds. No need for rest periods of off an on time.
5. On average the focused acoustics of Covaris uses 150 times less energy to shear your samples than a probe or water bath sonicators. That translates to a gentler shearing, that is more efficient, consistent, and controllable.
6. Optimized shearing settings for fragments in the size range of 100bp-5kb. These settings will work the first time. No need for optimization.
7. Have you checked a Bioruptor sheared sample directly on a bioAnalyzer without any clean up steps?

Thank you

Hamid

[epigeneticfan]

I think the counterattack should occur on two fronts:

1. I would argue against Hamid's claims of heating. He says that in the Bioruptor the samples heat up, causing AT rich regions melting, and being more susceptible to shearing, so the ends of the fragments (beginnings of the reads) are more likely to be AT rich. We should prove that this theory is untrue, the samples are in an isothermal environments, those heat maps only exist in Covaris brochure. Anyway, I don't think that if any heating occurs, it could reach the melting point of AT rich DNA fragments, and thus the spots of single stranded DNA breaks easier. If this could happen, it would be visible much more significantly.

2. we should provide real evidence of uniform shearing. I would emphasize that none of our customers have complained about this 'positional effect' (encourage forumites to come out if they have), nor have we seen it during our internal tests; instead last year a customer was happy to produce a whitepaper to show the uniform coverage he achieved with Bioruptor (link: http://www.diagenode.com/en/news/2011/11/case-study).
Unfortunately we only have one genomic DNA sequencing, but fortunately it has 76 bp long reads and it was sheared with both covaris and Bioruptor (it belongs to the mentioned whitepaper). Usually we make ChIPSeqs with short reads (36 bp).
I attached some files: the S. aureus (br = Bioruptor, cv = covaris) has a very AT rich genome, that's whiy it's graph is lower than E. coli's. You could show what you like, maybe you shouldn't show covaris data, our graph is enough. But as you can see, covaris data is almost the same, the shearing really doesn't make a difference. PCR, however, does, as you can see in the 'diffpcr' files: the only difference between them and the other E. coli samples is that they changed the PCR protocol there (but again, not much difference between Br and C, and no overall climbing of GC ratio).
I want to show only one file, I would choose S. aureus with Bioruptor, that graph is perfectly flat, no inclinations at all.

I also included a short read ChIPSeq file just for the sake of curiosity. Note the somewhat different scale (due to the shorter read length), but the GC content is very uniform across the whole length, the line almost perfectly flat here as well. It's a H3K9me3 ChIPSeq on human cells (I guess HeLa).

All in all - as others suggested - I am quite sure that the observed positional effect is due to the PCR, not the shearing.

We have a very friendly and helpful tech support team , and we will gladly give advices on shearing and enrichment if needed.