This article discusses some of the important things to consider if you are using agarose gel electrophoresis for size-selection of your NGS libraries. Gel electrophoresis is a simple and very commonly used technique in most labs.

Careful! It’s a critical step

However this simplicity means people can often overlook the fact that there are applications where the running of the gel is a critical step that must be carried out very carefully. Many of the NGS library preparation protocols use agarose gels for size-selection of genomic DNA, RNA or final libraries.

Don’t ignore the simple things

As a core lab running lots of next-gen sequencing we have seen many thousands of libraries. The early days were far from ideal due to the large variability in size. Much of this came from ignoring some simple things to consider in gel electrophoresis.

Four Top Tips to success

This article highlights some of the main issues; run your gels consistently and the analysis of your NGS data will be that little bit easier.

  1. Which agarose to use: It is not all the same! You can buy the basic and cheap agarose, or the more expensive (low melting point and high-resolution) variants. Each of these is suitable for a particular technique and using a different product for each gel will introduce differences in the final DNA selected. Many protocols specify a particular product and I’d always recommend you do exactly what the protocol says. The most important thing is to use the same agarose for all your experiments.
  2. How to make the gel: It is important that gels are made up using a standard operating procedure, as you need to produce the same percentage gel every time. Weighing everything carefully is key and it is worth weighing the final agarose:buffer mix before and after boiling to add back any water that has boiled off. This evaporation can change the concentration by a few percent. This may not sound a lot, but we are aiming for exactly the same results every time. You’ll probably be using either TBE or TAE buffers. TBE is less likely to overheat, but can inhibit downstream reactions. We use TAE for our NGS gels as the borate in the TBE can inhibit downstream reactions. However there are other buffers to use, a previous BiteSizeBio article explains the advantages of some newer buffers.
  3. How to run the gel: Always use the same conditions for running your gel. Time, voltage and temperature are the biggest factors in determining what size products you are likely to cut out at the end of your experiment.
  4. Getting your samples out of the gel: The easiest way to get your DNA out of the gel is to cut it out and use a gel extraction kit. However, simply using a scalpel is likely to result in different results each time as it can be difficult to cut exactly the same size from the gel. There are several suppliers of disposable gel cutting tips. These fit on the end of a pipette and result in identical gel cuts every time. Of course if you don’t follow steps 1, 2 & 3 carefully, then your results will vary every time!

Does not compute

Small variations can lead to significant differences in insert sizes. If the software you are using does not compute insert size from the sequence reads, then things like structural variation analysis or peak-shifting can be adversely affected.

If it goes horribly wrong…don’t panic!

One very useful tip is to cut out multiple bands. If something goes horribly wrong downstream you still have a sequencing library to go back to! However, by following the above tips, you’ll get the same size NGS library each time you prepare a new set of samples.


Originally published on May 2, 2013. Updated and Revised July 10, 2015.

More 'Genomics and Epigenetics' articles

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.