The Actual Process
You make a gel. Usually agarose, sometimes polyacrylamide if you need better resolution. Dissolve the powder in running buffer, microwave it until clear, pour it into a casting tray with a comb, and wait for it to set. That's the easy part. Then you load your samples mixed with loading dye into the wells, submerge the gel in a tank filled with the same buffer, run current through it, and visualize the results. The whole thing from start to finish takes roughly two hours if nothing goes wrong, which is rare on the first attempt. DNA is negatively charged because of its phosphate backbone. When you apply an electric field, the molecules move toward the positive electrode. The gel acts as a molecular sieve. Smaller fragments navigate through the pores more easily and travel faster. Larger ones get tangled up in the matrix and lag behind. That's why after staining, you see distinct bands at different positions depending on fragment size. RNA and proteins follow similar principles, though proteins need SDS to uniform their charge-to-mass ratio before separation becomes meaningful. I run gels probably three or four times a week in my lab work. The most common mistake I see people make is pouring gels that are too thick. A standard 5 to 7 mm gel thickness is fine, but when you cast them 10 mm or more, the bands smear because heat doesn't dissipate evenly and the current density varies across the depth. You lose resolution and waste sample. Always use a proper casting frame with consistent spacers.
Gel Concentration and What It Actually Does
The percentage of agarose determines pore size. A 1% gel separates fragments between 500 and 10,000 base pairs reasonably well. If you're working with smaller amplicons under 500 bp, switch to a 2% gel. For anything below 100 bp, you need polyacrylamide, usually 6 to 15% depending on the size range you care about. The relationship isn't linear though. Going from 1% to 2% doesn't halve the pore size in a simple way. It's more complex because the polymer network forms a random mesh, and the effective pore radius scales with the inverse square root of the percentage approximately. Here's something people miss: higher agarose percentages don't just slow everything down. They also increase resistance, which means more heat is generated at the same voltage. If you run a 2% gel at 120 volts the way you'd run a 1% gel, the gel can literally melt or crack. Drop the voltage to 80 to 100 for higher percentage gels and extend the run time. It gives sharper bands and saves you from scraping melted plastic out of your tank.
The Buffer System Matters More Than You Think
TAE and TBE are the two standard options. TAE has lower buffering capacity, which means it can't handle high voltages or long runs without pH shifting. You'll see band smearing if you run a TAE gel for more than an hour at high voltage. TBE holds pH better and gives sharper bands for small fragments, but the borate can interfere with downstream applications like cloning or sequencing if you don't recover the DNA properly. If I'm running a diagnostic gel just to check a PCR product, TAE is fine. If I'm preparing bands for extraction and ligation, TBE and then a cleanup step afterward is the safer route. I once spent a full day troubleshooting why my restriction digests kept showing extra bands on a 1% TAE gel run at 150 volts for 90 minutes. The bands weren't real. They were artifact bands caused by the buffer running out of capacity mid-run. The pH dropped, the DNA secondary structure partially relaxed, and you got weird migration patterns. Switched to fresh TBE and ran at 100 volts for 45 minutes. Clean bands immediately. Always make fresh running buffer and never reuse it more than a couple of times. The ions get depleted and the conductivity changes unpredictably.
Get the Full Details
Loading, Running, and Staining
Mix your DNA sample with loading dye. The dye has glycerol or sucrose to make the sample sink into the well, and tracking dyes like bromophenol blue and xylene cyanol that move at predictable rates through the gel. Don't overfill the wells. Overflow causes smearing into adjacent lanes. A microliter or two past the well rim is enough. Run voltage depends on gel length. The rule of thumb is about 5 to 8 volts per centimeter of gel. A standard mini-gel that's roughly 10 cm long runs well at 80 to 100 volts. Higher voltages generate more heat and broaden bands through thermal convection. Lower voltages give cleaner separation but take longer. A 1% gel with a 500 bp to 3 kb ladder usually resolves in 35 to 50 minutes at 90 volts. For visualization, ethidium bromide is the cheap standard. You can either embed it in the gel before pouring or stain the finished gel afterward. Post-staining takes longer, usually 20 to 30 minutes with gentle shaking, but it's safer for the gel structure since the casting process doesn't involve handling the mutagen. SYBR Safe and GelRed are commercial alternatives that are less mutagenic though not completely harmless. Silver staining for polyacrylamide gels is another option when you need femtogram-level sensitivity, but it's fussy and the protocol takes about two hours.
Common Problems and What Actually Fixes Them
Fuzzy or smearing bands usually mean one of three things: the gel was too warm during the run, the sample was overloaded, or the DNA was degraded. Check your voltage. If the gel feels warm to the touch after 20 minutes, you're running too hot. Spread the sample less. A typical load is 100 to 500 nanograms of DNA per lane for agarose. More than that and the bands broaden from overloading the dye binding capacity and the physical space in the pore matrix. Another issue that comes up frequently: DNA stuck in the well. This happens when the sample density is wrong, the well integrity is compromised, or the DNA is too large for the gel percentage. I had a client once trying to separate a 15 kb PCR product on a 2% gel. The DNA barely entered the gel. It just sat in the well and degraded over the run time. Switched to a 0.8% gel and the band resolved clearly in about 90 minutes at 90 volts. Always match your gel percentage to your expected fragment size range. Bands running diagonally or skewed to one side usually indicate uneven gel thickness or an electrode connection problem. Check that the leads are making solid contact and that the gel surface is level. A tilted tank causes the current path to be asymmetric and the migration pattern follows.
When Gel Electrophoresis Isn't the Right Tool
This method has real limitations. Resolution tops out around 1 to 2% for agarose. You can't reliably distinguish fragments that differ by less than 5 to 10% in size, which means two products that are 500 and 520 base pairs will run as a single broadened band on a standard agarose gel. Capillary electrophoresis or denaturing gradient gel electrophoresis handles that kind of resolution, but those require specialized equipment. For routine clone checking, PCR validation, or restriction digest verification, agarose gel electrophoresis is still the workhorse. It's fast, cheap, and gives you enough information to make a decision. Just don't expect it to solve problems that are outside its resolution range. The other hard limitation is quantification. You can estimate concentration by comparing band intensity to a known ladder, but it's rough. Maybe 20 to 30% accuracy at best. If you need precise quantification, use a spectrophotometer or fluorometer instead. Gel electrophoresis tells you size and approximate amount. That's it. It won't replace a Nanodrop or Qubit when you're preparing samples for library construction or next-generation sequencing.
