ELISA: How to Analyze Experimental
An ELISA test must always be performed in replicates (doubles or triplicates) in order to obtain a sufficient number of samples to calculate the standard deviation and the coefficient of variation. The error coefficient must be kept below 20%. If the calculated error is high, errors due to pipetting, contamination of plates and reagents, evaporation and/or temperature should be evaluated.
A standard curve is used to measure the concentration of the sample. First, the absorbance values of the standard concentrations (pg/ml) are plotted. A negative curve is also performed which contains no standard sample data and is used to detect and measure background noise. For an accurate measurement, this background value must be subtracted from the standard curve.
Absorbance values of known concentration can be plotted as a positive control. The target protein concentration is derived by first calculating the average absorbance and then using the standard curve to estimate the concentration.
Dilution factors must also be taken into account. For example, if the sample was diluted four times, the absorbance must be multiplied by four before using the standard curve to derive the concentration.
The definition of the coefficient of variation is the ratio of the standard deviation and the mean, i.e. standard deviation (σ)/mean (µ). High values of the coefficient of variation indicate inaccuracies due to pipetting or mixing of reagents between wells, bacterial or fungal contamination, dry wells, or temperature variations.
The value obtained after running the ELISA depends on how a targeted protein interacts with the antibodies, and this interaction is compared to a standard protein curve. Several factors – including buffer, matrix, and heterophile antibodies – can affect sample binding, so these should be considered when determining the accuracy of an ELISA.
A peak test can be performed to determine the accuracy of ELISA results. In this test, a known amount of recombinant protein is added or added to the sample. Then, using the standard curve, the amount of material is measured. Large deviations in measured values may indicate errors due to unknown elements in the assay.
The linearity test is also used to determine the accuracy of ELISA. In this case, the "spiked sample" is serially diluted at 1:2, 1:4, and 1:8. If the changed sample concentration is detected and the results do not show a linear change in the measurements, this may indicate errors in the precision of the ELISA.
In some cases, errors can be observed at certain concentrations and no longer be observed at lower concentrations. In such cases, the sample may need to be diluted before concentration measurements are made.