[vote] The best Gel extraction kit - which brand do you think is the best ? (Aug/07/2009 )

I tried Roche, promega, Bioneer, Qiagen. I would say, Roche kit give the best yield


how about you guy ?

Best approach: design your protocol to eliminate it.

My favorite is GFX kit from GE Healthcar
.
Can't tell you how it goes with the yield comparing to other kits, but it's clear advantage is that it requires three spins instead of five: Only one spin to load (you add equal volume of capture buffer to dissolve agarose, not several volumes) and one spin with ethanol (again, only 500 ul of ethanol which is removed completely without additional centrifuging step)

Hey phage434, your idea is interesting, but how do you eliminate gel purification?

Hi frens!

For my work, Roche kit gave low yield! I suggest INTRON kit (for high yield) and GeneAll (for high purity). Check this out!

Cheers,
appleyun

I've used Qiagen (minelute and qiaquick), an invitrogen kit, and a zymo kit.

Qiagen and invitrogen seemed pretty comparable. While using the zymo kit, a purification would yield no detectable DNA, but using the same conditions with Qiagen gave a pretty good amount, so zymo seemed pretty poor for me.

For cloning, it's not so much the quantity of DNA as it is the quantity of undamaged DNA. You can improve your cloning results dramatically by following the simple step of adding 0.28 g/L guanosine to 1x TAE as a UV protectant (stir with gentle heat until dissolved), and using this to cast and run your gel. See here.

We use the Qiagen kit, and have had good results with it, but I think the results are influenced more by adding the guanosine than by the particular kit used.

I like Qiagen and Zymo.

Run two lane and the same time. split to half, one get EtBr and one doesn't.
use that as reference.

no UV yay!

hanming86 on Aug 12 2009, 09:46 AM said:

That requires twice as much sample. Just make up 4L of 1x TAE + gaunosine for casting and running gels from which you're going to recover bands and be done with it...

HomeBrew on Aug 13 2009, 06:08 AM said:

I used to do the method hamming86 mentioned, don't need much of sample, just some ul of extra sample to use as a control. For example: run 100 ul sample on the left lane + 10 ul sample on the right lane. After running, cut the gel into 2 parts, stain only the right one in EtBr, mark the location of the band by observing on UV lamp and cutting. Then place the two parts together, excise the large, unstained one based on the marker sites on the small one. Confirm the cut by staining the remaining gel. No EtBr, no UV ^^