I realize this is an old question, but I just saw it before trying an Ampure XP purification on genomic DNA myself, so I'll contribute my 2 science cents. Happy to report that it worked, I did a 2x purification (50 uL sample + 100 uL beads) and I'm not seeing any significant sample loss judging from a quick gel. The gDNA is from Aspergillus nidulans.

Removing the buffer by vacuum centrifugation or drying the tubes with caps off in 50 degree C incubator to remove buffer and add 10mM Tris buffer is an alternative. In theory you will not loose DNA this way.

Thank you for your answer. We really want to use magnetic beads since we have them in the lab and they are convenient. Our material is very precious,we don't want to lose it and want to proceed fast.

I have used the AMPure XP system on PCR amplified captured DNA for sequencing. Works great with the magnetic ring plats. Typically the DNA fragments are between 200-600 bp in size. Whether or not it works with gDNA needs to be tested. "Genomic" DNA is typically fragmented from >5kb up to tens of kb fragments and the beads are designed to capture DNA >100bp according to the manual. No info for gDNA. You might want to test with a control gDNA of good quality first and assess the yield before/after the capture and elution. For sure the yield will not be 100%.  In our lab we use evaporation for concentrating precious DNA samples in low concentrations.

If your sample is precious I would precipitate with ethanol, vacuum drying is also good suggestion. Then wash with 70% EtOH. If you insist to use AmpureXP - yes it will work up to 0.6 : 1 , the tricky part is to not let the beads completely dry after 70% EtOH wash. Ampure is based on electrostatic binding of DNA to COOH groups on the surface, when the surface get completely dry this binding becomes hydrophobic  which is much harder to break and that will result in significant loss of your precious sample.

my2c

From experience you can use Ampure XP to purify genomic DNA (at least from bacteria).  We use 1.5 volumes of XP for 1 volume of sample.  I'm not sure of the yield but the output is certainly good enough for NexteraXT library preparation.

Alternatively the input requirement for NexteraXT is so low that you can probably just dilute your sample with water, thus reducing the EDTA concentration.

I am using the Nextera XT kit to prepare my librairy. You Can use the Angencourt Ampure XT to purify your gDNA. If You want to run on 2x250 or 2x300bp, You must use 25microliters of beads by sample before the librairy normalisation. Those beads are able to capture nucleic acids with more than 100bp and leaving contaminants in solution like enzymes, primers,...The ampure beads are more concentrated than beads That are used to normalise the librairy. Do not hesitate if u have  others questions...

Thank you for your answers. I agree with you Richard, it would have been better to simply dilute our samples (only 2.5 ng/µL is needed), but the problem is that the concentrations of our gDNAs are so low that it's not possible, we actually also need to concentrate them! That was our idea behind: getting rid of EDTA because we can not simply dilute them (concentrations ranging frome 3 to 15 ng/µL).

I once performed a quick test by purifying 7 ug of Drosophila genomic DNA by either of these methods:

1) vaccumn drying  and resuspending in water => got 6.4 ug (91%) of DNA back

2) AMPure bead purification (used 90ul of beads for 50ul of DNA solution) and elution in water => got 4.1 ug (58%) of DNA back

So, out of those two, I would recommend vacuumn drying, at least for these amounts of DNA. (But it could be that AMPure beads' binding capacity is not enough to capture this much of DNA and they will give better yield with lower amounts of input DNA).

We have very low amounts of gDNA and I compared EtOH precipitation, isopropanol precipitation and ampure XP beads. I got a much higher recovery rate with magnetic beads (around 80%) compared to the two other methods (something like only 20%). But I work with concentrations of 5 ng/µL approximately.

I'm reading the answers here with a bit of amusement or at least confusion, since some people suggest vacuum drying as a way of removing EDTA or Tris. Please, could you confirm that EDTA or Tris are volatile subtances? :)

Works for me. Just give it a go.

I would suggest against it, Ampure XP beads have a "max" bp range of 10k or so.  You will probably recover some genomic DNA anyway, but the yield won't be great.Maybe try an ethanol or isopropanol precipitation and then elute in molecular grade water?  It takes longer and involves more prep but probably a higher yield.

Zymo Research has DNA clean and concentrate kits if you like spin columns. It works on genomic DNA.

Vaccum drying won't make EDTA evaporate. EDTA isn't as volatile as EtOH :)

I wouldn't do AMPure bead clean-up. There is always a 30%-50% sample loss with bead clean-up. Clean and concentrate your sample with a column that has a higher binding capacity for HMW DNA. Columns do have limitations when comes to the length (bp) and amount (ng or ug) of genomic DNA.

Noemie, what method did you finally decided to use and what were the results? Would you recommend Ampure for genomic DNA purification?

I ended up having success using either Ampure or simple ethanol precipitation with the latter being less fiddly.

I finally used magnetic beads for all my gDNA samples (1X volume) and it appeared to work perfectly fine. Especially with those cheaper beads called CleanNGS (provided by Clean NA). I got them for half the price of the Agencourt beads, and they work for both DNA and RNA.

Thanks for all your advices. 

That's great! Thanks for the feedback.

I realize this is an old question, but I just saw it before trying an Ampure XP purification on genomic DNA myself, so I'll contribute my 2 science cents. Happy to report that it worked, I did a 2x purification (50 uL sample + 100 uL beads) and I'm not seeing any significant sample loss judging from a quick gel. The gDNA is from Aspergillus nidulans.

You can even use less, actually. In my case, it worked perfectly for gDNA purification before my sequencing library prep, but it was also the opportunity to concentrate some of these samples.

The only problem I am facing is the quantification. I am using Qubit (as recommended by Illumina), but concentrations vary a lot from one measure to the other. I guess it's a problem when working with low concentrations of gDNA: if you pipet "big noodles" or not, it will make a lot of difference.

So I did intermediate dilutions: concentrated gDNA (mseasured) > diluted to 0.5 ng/µL (measured) > diluted to 0.2 ng/µL (not measured)

Hope this helps for others...

SPRI beads work fine on practically any DNA. They should not have a "max" DNA bp range, but they do have optimal range for the template DNA concentration. This range should be tested by simply making a dilution gradient with your DNA. This takes only a few hours hands-on time.

Please check for example our latest paper where we purified very hard-to-work-with Fungi using our own SPRI bead solutions: Article Finding flies in the mushroom soup: Host specificity of fung...

The instructions on how to make very cheap and easy SPRI beads: Article What you need is what you eat? Prey selection by the bat Myo...