Question 6 of 6 Advanced Concept
How can normalization obscure a failing assay control?
Short answer
Scaling each plate to its own controls can make a degraded control appear acceptable. Always inspect and apply run-quality limits to raw controls before interpreting normalized values.
Why
Consider a plate where the positive control falls from 1000 to 500 signal units because a reagent degraded, while the negative stays near 100. Normalizing the positive to 100% masks that absolute loss and can make samples look comparable across plates. It may also amplify noise when the reference dynamic range narrows.
Plot raw distributions, control separation, drift and Z-prime for each run. Reject plates that fail predefined checks rather than forcing them onto a common range. Maintain common calibrators or bridge samples to identify across-run shifts.
Then calculate normalized values only for plates meeting technical acceptance criteria and disclose the transformation. Spatial normalization can improve known plate effects, but it should not erase real biology or replace troubleshooting. If treated samples affect the normalization references, choose an appropriate independent control.
The goal is interpretable data, not numerical uniformity at any cost.
Common mistakes
- Interpreting a normalized 100% positive control as proof that the actual raw positive control passed QC
The interviewer might ask next
What diagnostic plot would show that a seemingly normal normalized plate had lost its assay window?