Assay QC & Validation
Z'-factor, CV, signal window, outlier detection, and validation metrics for your assay
Before you trust a number from an assay, you should trust the assay. This workflow grades plate quality, quantifies precision, validates a method, and flags outliers - the diagnostics that decide whether a run is worth analyzing at all.
What it measures
This bundles three related checks:
- Assay QC. Z'-factor and signal window from your positive and negative controls - the headline numbers for whether a screen can separate hits from background.
- Assay validation. Precision (intra- and inter-plate CV), accuracy/recovery, linearity, and limits of detection and quantification.
- Outlier test. Grubbs' or ROUT detection on a set of replicates, so a single bad well is caught rather than averaged in.
Data you upload
- Control wells (positive and negative) for Z'-factor and signal window.
- Replicate measurements for CV, and a dilution/recovery series for accuracy and linearity.
- For the outlier test, the replicate group you want screened.
What you get back
- Z'-factor with the standard interpretation band (> 0.5 is an excellent assay; 0-0.5 is marginal; below 0 is unusable).
- %CV for each replicate group, signal-to-background, and the signal window.
- LOD/LOQ and a recovery/linearity summary where applicable.
- Any flagged outliers, with the statistic and significance that flagged them.
Method note
The Z'-factor combines the means and standard deviations of both controls into a single dimensionless score, so it rewards both a wide signal window and tight replicates. Outlier detection defaults to Grubbs' test for single outliers and ROUT for larger sets; nothing is removed automatically - flagged points are surfaced for you to confirm, because silent removal is a reproducibility hazard.
Related
Per-fit quality flags for a specific workflow live in Quality Control. For the precision of a measurement method against a reference, see Bland-Altman. The underlying formulas are in Statistical Methods.