Skip to content
日本語Newsletter

Can this assay tell hits from non-hits?

Z-factor (Z′) calculator for assay quality

Takes the means and standard deviations of your positive and negative controls and sums up, in one number, how cleanly the two signals separate.

When to use itUse it before a high-throughput screen (HTS) of thousands of compounds, to decide whether the assay is ready for the real run.

Prefilled with an example: a plate with a positive-control mean of 10,000 and a negative-control mean of 1,000. Change any value and the result updates.

Mean of the wells that always respond (positive control). Any signal unit works, such as fluorescence intensity, as long as all four fields use the same one.

How widely the positive-control values spread, as a standard deviation.

Mean of the wells with no response (negative control, background).

Standard deviation of the negative control.

Z′-factor (separation)0.7333
Signal-to-background ratio (S/B)10
Positive-control CV5 %
Negative-control CV30 %

The positive and negative signals separate clearly, and the assay is good enough for screening (at 0.7 or above, with headroom to spare).

How to read it

The row that matches your result is marked.

  • Z′ = 1Ideal (no variation)The theoretical maximum. Real measurements almost never reach it.
  • 0.5 to below 1Your resultWell separatedGood enough for screening, with more headroom at 0.7 or above. A Z′ of 0.5 or more is the usual benchmark for passing.
  • Above 0, below 0.5Usable, but worth improvingUsable for a pilot run, but the two distributions are close. Depending on where you draw the hit threshold, you will miss more true hits or pick up more false ones.
  • 0 or belowNo separationPositives and negatives overlap. The assay is not fit for screening.

How it is calculated

The closer Z′ is to 1, the more cleanly positives and negatives separate. The 3 in the formula comes from looking three standard deviations either side of the mean, a band that holds about 99.7% of normally distributed data. The calculator also gives the signal-to-background ratio (S/B = μp ÷ μn) and the spread relative to the mean (CV) for reference.

Z′ = 1 − 3 × (σp + σn) ÷ |μp − μn|

Source: The Z′-factor as defined by Zhang JH, Chung TDY and Oldenburg KR (J Biomol Screen, 1999).

With your numbers

  1. Z′ = 1 − 3(σp + σn) ÷ |μp − μn| = 1 − 3(500 + 300) ÷ |10,000 − 1,000| = 0.7333
  2. S/B = μp ÷ μn = 10,000 ÷ 1,000 = 10

Limits of this calculation

  • The thresholds are guides. What counts as acceptable depends on the assay type and detection method, and 0.5 is not an absolute pass line.
  • Cell-based and image-based assays vary more than biochemical assays with purified enzymes. Some labs run them at around 0.4 and refine how hits are called instead.
  • If outliers distort the result, a robust Z′ uses the median instead of the mean and the median absolute deviation (MAD) instead of the standard deviation.
  • Z′ describes the assay itself, from controls only. The Z-factor, which uses the distribution of real samples, describes performance in the actual screen and serves a different purpose.

Primary sources

Every calculator gives an estimate for learning, built on published information and primary sources. Real decisions follow the applicable guidance and expert judgment.