Transformation efficiency calculator
CFU per µg of DNA from colony count, DNA amount, recovery volume and volume plated.
CFU per µg of DNA from colony count, DNA amount, recovery volume and volume plated.
Transformation efficiency is the number of transformed colonies obtained per microgram of plasmid DNA, written as CFU/µg or transformants/µg. It is the standard measure of how competent a batch of cells is, and it lets you compare home-made cells with commercial ones, heat shock with electroporation, or today's batch with last month's. Nobody transforms a whole microgram; the value is scaled up from the picograms that actually reached the plate.
The total volume is everything in the tube when the plating aliquot is taken: cells, DNA and recovery medium. The dilution is 1 if you plated straight from the recovery culture, 10 if you first diluted it 1:10, and so on.
Worked example. 100 pg of pUC19 is added to 50 µL of competent cells. After heat shock, 950 µL of SOC is added (total 1000 µL). After one hour the culture is diluted 1:10 and 100 µL is plated, giving 150 colonies. DNA plated = 100 pg × 100 ÷ 1000 ÷ 10 = 1 pg = 10⁻⁶ µg. Efficiency = 150 ÷ 10⁻⁶ = 1.50 × 10⁸ CFU/µg. The whole reaction contained 150 × 10 × 10 = 15,000 transformants.
| CFU/µg (pUC19) | Typical source and use |
|---|---|
| below 10⁶ | Tired or badly prepared cells. Enough to retransform a purified plasmid, nothing more. |
| 10⁶ – 10⁷ | Home-made CaCl₂ cells. Plasmid retransformation and easy, high-yield ligations. |
| 10⁷ – 10⁸ | Good home-made (Inoue, TSS) or subcloning-grade commercial cells. Routine cloning. |
| 10⁸ – 10⁹ | High-efficiency commercial chemically competent cells. Blunt ligations, Gibson and Golden Gate assemblies, large plasmids. |
| above 10⁹ | Electrocompetent and premium chemical cells. Libraries and very low DNA input. |
Multi-fragment assemblies lose colonies with every added part, so check the expected yield of the assembly itself with the Gibson assembly calculatorbefore blaming the cells.
The usual causes are cells that were warmed during preparation or storage (each freeze–thaw costs about half the efficiency), a heat shock that was too long or too short for the tube type, skipping the recovery hour for antibiotics other than ampicillin, too much DNA or ligation mix (keep it under 10% of the cell volume, since ligase buffer and PEG inhibit transformation), and impure DNA carrying phenol, ethanol or detergent. For electroporation, salt in the DNA causes arcing; desalt or dilute the ligation first.
Divide the number of colonies by the micrograms of DNA that ended up on the plate. DNA on the plate is the DNA added to the cells, times the fraction of the recovery culture that was plated, divided by any dilution. For 150 colonies from 100 pg of pUC19 in 1 mL, diluted 1:10 with 100 µL plated: DNA plated is 0.001 ng, or 10^-6 µg, so the efficiency is 150 / 10^-6 = 1.5 × 10^8 CFU/µg.
Home-made calcium chloride cells give 10^6 to 10^7 CFU/µg, which is enough for retransforming plasmids. Commercial chemically competent cells give 10^8 to 10^9, and electrocompetent cells 10^9 to 10^10. Routine ligations work at 10^7 and above; libraries, Gibson assemblies with many fragments and large constructs want 10^8 to 10^9 or more.
Efficiency is only comparable when the DNA is a small, supercoiled plasmid at a non-saturating amount. Above roughly 1 to 10 ng per 50 µL of cells the colony number stops rising in proportion to the DNA, so the calculated CFU/µg falls. Ligation mixtures and large plasmids transform far less efficiently than pUC19, so they measure the DNA, not the cells.
The volume of the cell suspension at the moment you take the aliquot for plating: competent cells plus DNA plus the SOC or LB added for recovery. With 50 µL of cells and 950 µL of SOC it is 1000 µL. If you spin the cells down and plate everything, set the plated volume equal to the total volume.
Use the nanograms of vector DNA in the volume of ligation mix that you added to the cells, not the whole ligation. A 20 µL ligation with 50 ng of vector, of which 2 µL was transformed, contributes 5 ng. The result describes the ligation and the cells together, and is normally 10 to 1000 times lower than the pUC19 control.