ELISA

How to analyse ELISA data: 4PL, 5PL and weighting

Turn plate-reader absorbances into reliable concentrations: blank correction, the right curve model, weighting and quality checks.

1. Correct for the blank

Average the blank wells (zero standard or assay buffer) and subtract that value from every standard and sample. Then average technical replicates; a coefficient of variation (CV) above about 15–20% between replicates is worth a second look before you trust the value.

2. Choose the curve model

Sandwich and competitive immunoassays give S-shaped curves when concentration is on a log scale, so a straight line through all standards is rarely appropriate.

• 4-parameter logistic (4PL): y = d + (a − d) / (1 + (x/c)^b), where a and d are the lower and upper plateaus, c is the inflection point (EC50) and b the slope. The standard choice for most kits.
• 5-parameter logistic (5PL): adds an asymmetry parameter g. Use it when the curve is clearly not symmetric around the midpoint and you have enough standards (at least 6–7 levels).
• Linear: only for the straight middle part of a curve, or for kits whose manual asks for it.

3. Weight the fit

The spread of OD readings grows with the signal, so unweighted fits let the top standards dominate and make low concentrations inaccurate. Weighting by 1/Y² gives each standard a similar relative influence and usually improves accuracy at the low end, where many biological samples fall.

4. Check the curve before reading samples

• Back-calculate each standard from the fitted curve; recovery of roughly 80–120% of the nominal value is a common acceptance range.
• R² close to 1 is necessary but not sufficient: look at recovery and at the residuals.
• Do not extrapolate: report samples above the top standard as “above range” and re-run them at a higher dilution; samples below the lowest standard as “below the limit of quantification”.

5. Apply the dilution factor and units

Multiply interpolated values by the dilution factor. If you need molar units, convert with the protein’s molecular weight (for example pg/mL ÷ kDa = pM).

6. Compare groups

Concentrations are often skewed; inspect the distribution and consider a log transformation or a non-parametric test. The test guide helps you choose.

Published 2026 by Dr. Ambuj Shahi. This guide is general information for researchers; follow your institution’s and journal’s requirements.