Protein concentration assay
Paste standards and sample absorbances. Fits the standard curve and reads protein concentrations from it.
Paste standards and sample absorbances. Fits the standard curve and reads protein concentrations from it.
Colorimetric protein assays do not give a concentration directly. Each one turns protein into colour, and the absorbance is compared with a set of standards of known concentration, usually BSA or bovine gamma globulin, run on the same plate. The standards define a standard curve; the unknowns are read from it. This calculator fits the curve by least squares, reports the equation and R², and converts every unknown absorbance to a concentration, multiplied by the dilution factor of that sample.
Worked example. BSA standards at 0, 125, 250, 500, 750 and 1000 µg/mL read 0.052, 0.166, 0.291, 0.528, 0.757 and 0.998 at 595 nm. The linear fit is A = 0.000946 × c + 0.0517 with R² = 0.9999. A lysate diluted 1:10 reads 0.612, so it contains (0.612 − 0.0517) ÷ 0.000946 = 592.6 µg/mL in the assay (with unrounded coefficients) and 5926 µg/mL, about 5.9 mg/mL, undiluted. The example link above the result loads these numbers.
| Assay | Read at | Working range and interference |
|---|---|---|
| Bradford | 595 nm | About 100–1500 µg/mL (1–25 µg/mL micro format). Fast, tolerant of reducing agents, but disturbed by detergents such as SDS and Triton, and the response varies between proteins. |
| BCA | 562 nm | About 20–2000 µg/mL. Tolerates most detergents up to 5%, but reducing agents (DTT, β-mercaptoethanol) and copper chelators (EDTA) interfere. |
| Lowry | 750 nm | About 5–2000 µg/mL depending on format. Sensitive, but slow and disturbed by many buffer components including Tris, detergents and reducing agents. |
The arithmetic is the same for all three, and for any other assay with a standard curve, such as the biuret or a fluorescent dye assay. A pure protein with a known extinction coefficient needs no standard curve at all: measure A280 and use theA280 protein concentration calculator.
Over a narrow range the response is a straight line. Over the full kit range Bradford and BCA curves bend: the signal per microgram falls as the reagent is used up. A straight line forced through a bent curve overestimates the middle and underestimates the top, which shows in the recovery column as standards that back-calculate far from their nominal value.
The quadratic fit adds one curvature term and is what most plate reader software offers as "second-order polynomial". The calculator takes the root on the rising side of the parabola. It needs at least three standards, and four or more for R² to mean anything. Do not read samples beyond the highest standard with either model; a polynomial is only a description of the curve inside the measured range. Sigmoidal immunoassay curves need a four-parameter logistic instead, which the ELISA data analyzer fits.
Standards: one per line, concentration then absorbance, separated by spaces, tabs or commas, so two columns copied from a spreadsheet paste straight in. Replicate readings can follow as extra columns and are averaged. Include the zero standard. Results come back in the units of the standards.
Unknowns: one per line: a name, the absorbance, and an optional dilution factor. A sample diluted 1:10 before the assay has a dilution factor of 10. If standards and samples were not blanked by the reader, enter the blank absorbance under the curve and it is subtracted from every reading. To plan the sample dilution itself, use thedilution calculator.
Fit a straight line through the absorbance of the standards against their concentration, A = slope × concentration + intercept. For each unknown, subtract the intercept from its absorbance, divide by the slope, and multiply by the dilution factor. With a slope of 0.000946 per µg/mL and an intercept of 0.0517, an absorbance of 0.612 is (0.612 − 0.0517) ÷ 0.000946 = 592 µg/mL in the well.
Use a linear fit when the standards fall on a straight line, typically R² of 0.99 or better over a narrow range. Bradford and BCA curves bend at higher concentrations because the dye or the copper chelate approaches saturation; if the line misses the top standards, a quadratic (second-order polynomial) fit follows the curvature and gives better back-calculated values.
Most labs accept R² of 0.99 or higher for a linear fit of six to eight standards. A lower value with a smooth, bending curve calls for a quadratic fit or a narrower standard range. A lower value with one outlying point calls for repeating that standard.
The concentration is extrapolated and unreliable, because the curve flattens beyond the standards. Dilute the sample so that it reads in the middle of the curve, measure it again and enter the dilution factor.
Either blank the instrument on the zero standard, subtract the blank reading from every well, or include the zero standard as a point in the curve. All three are equivalent as long as standards and samples are treated the same way. The intercept of the fitted line absorbs a constant background.