Z-factor calculator

What is the Z-factor?

The Z-factor, introduced by Zhang, Chung and Oldenburg in 1999, is a single number that grades how well an assay separates its positive and negative controls. It combines the distance between the two control means with the spread of their replicates, so a large signal window with tight replicates scores high and a small or noisy window scores low. It is the standard quality metric for high-throughput screening (HTS) and for validating any plate-based assay.

Z-factor formula

μ₊ and σ₊ are the mean and standard deviation of the positive control, μ₋ and σ₋ those of the negative control. The numerator is the width of the two 3-SD bands and the denominator is the signal window. Standard deviations use the sample formula (n − 1). Z′ can never exceed 1 and has no lower bound.

Worked example. Positive control mean 96.0 with SD 1.0; negative control mean 12.0 with SD 0.7. Z′ = 1 − 3 × (1.0 + 0.7) / 84.0 = 1 − 0.061 = 0.94, an excellent assay.

How to interpret Z′

Z′ versus Z-factor and signal-to-background

• Z′ (Z-prime) uses the two controls only and measures the assay. Report it for every plate during a screen.
• Z-factor replaces the positive control with the screened samples, so it also reflects how active the library is.
• Signal-to-background (S/B), μ₊/μ₋, and the signal window, μ₊ − μ₋, describe the dynamic range but ignore variability. A high S/B with noisy replicates can still give a poor Z′, which is why Z′ is preferred.
• CV % (σ/μ × 100) reports the relative variability of each control on its own.

How to improve a low Z′

1. Widen the signal window: optimise reagent concentrations, incubation time and cell number so the positive control is as far from the negative as the readout allows.
2. Reduce replicate variability: check pipetting, edge effects, evaporation, mixing and reader settings. Well-to-well CV under 10 % is a good target.
3. Increase replicates on validation plates, 16 or more per control, so the estimate of Z′ itself is stable.

Related tools: the ELISA data analyzer for 4PL standard curves and the cell viability calculator for viability readouts that feed into Z′.

Frequently asked questions

What is the difference between Z-factor and Z′ (Z-prime)?

Both use the same formula. Z′ is computed from the positive and negative controls only and describes the quality of the assay itself. Z-factor is computed from the screened samples versus the negative control and also reflects the spread of the compound library. This calculator gives Z′ from control replicates.

What is a good Z′ value?

Z′ above 0.5 is an excellent assay suitable for high-throughput screening. Between 0 and 0.5 the assay is marginal: hits can be found but with more false calls. Below 0 the control distributions overlap and the assay is not usable. A Z′ of 1 is the theoretical ideal with zero variation.

How many control replicates do I need?

At least two per control are required to compute a standard deviation, but Z′ from so few values is unreliable. Typical validation plates use 16 to 32 wells of each control, and Z′ should be checked on every screening plate.

Why is my Z′ negative?

A negative Z′ means the 3-SD bands of the two controls overlap: 3(σ+ + σ−) is larger than the separation |μ+ − μ−|. Either the signal window is too small or the replicate variability is too high. Increase the dynamic range, reduce pipetting or edge effects, or change the readout.

Can I use Z′ for cell-based or ELISA assays?

Yes. Z′ is readout-agnostic: use viability, absorbance, fluorescence or luminescence values directly. For ELISA standard curves see the ELISA data analyzer, and for viability readouts the cell viability calculator on this site.