Understanding Z′-factor in high-throughput screening

The single most important quality metric for your assay: what it measures, how to calculate it, and what to do when it fails

What Z′-factor measures

Z′-factor (Z-prime) is a dimensionless score between −∞ and 1 that summarises the separation between your positive and negative controls relative to their combined variability. It was introduced by Zhang et al. (1999) specifically for HTS assay validation.

A Z′ of 1 is a perfect assay, with no variability whatsoever. In practice, anything above 0.5 is considered screening-ready.

Z′ valueAssay qualityRecommendation
> 0.6ExcellentProceed to full screen
0.5 – 0.6GoodAcceptable; monitor plate by plate
0 – 0.5MarginalOptimise assay before screening
< 0FailedPositive and negative controls overlap; do not screen

The formula

Z′ is calculated from the means and standard deviations of your positive and negative controls on the same plate:

VariableMeaning
μ₊, μ₋Mean of positive and negative controls
σ₊, σ₋Standard deviation of positive and negative controls
Z′ = 1 − (3σ₊ + 3σ₋) / |μ₊ − μ₋|

The numerator captures the combined spread (three standard deviations on each side, a 99.7% confidence envelope). The denominator is the signal window.

Widen either control's spread or pull the means together and watch the 3σ bands eat the signal window, and Z′ collapse:

densityneg. controlpos. controlsignal window

Z′ = 0.40: marginal: usable but noisy. Z′ rewards a wide gap between controls and punishes their spread: widen either σ or shrink the separation and the 3σ bands eat the signal window. Aim for Z′ ≥ 0.5.

Calculate Z′ per plate, not as a single value across your entire run. A screen with an average Z′ of 0.6 can still contain individual plates with Z′ below 0.3 that would generate unacceptable false-positive rates.

Common causes of low Z′

CauseSymptomFix
Small signal windowμ₊ and μ₋ are close togetherOptimise compound concentration, cell density, or incubation time
High control variabilityσ₊ or σ₋ is largeCheck dispensing precision; inspect for well-to-well position effects
Outlier control wellsOne well far from the othersFlag and exclude; investigate pipetting or evaporation
Reagent degradationDecline in Z′ across a multi-day runPrepare fresh reagent batches; store at correct temperature

Z′ versus Z-factor

Z′-factor uses only controls. Z-factor (no prime) uses the actual sample distribution versus one control group. For assay validation, always report Z′. Z-factor is occasionally used post-screen to assess hit identification performance.

Never average Z′ across plates to report a single screen-level value. Report the distribution: median, minimum, and the number of plates below threshold. Averaged Z′ conceals plate-level failures.

References

  1. Zhang, J. H., Chung, T. D. Y., & Oldenburg, K. R. (1999). A simple statistical parameter for use in evaluation and validation of high throughput screening assays. Journal of Biomolecular Screening, 4(2), 67–73.

Ready to run this analysis on your own data?