Skip to content

Use it for · Plate and instrument QC

Does every well of your cycler give the same Cq?

Run one reaction in every well and import the file. Clarida judges the spread, names each stray well and shows on a heat map where it sits.

4 in your first month, then 2 a month. No card.

New experimentqPCR start page › Plate homogeneity check
Already ran the plateNot here: the test plate is set up first
On this page

What you get

A verdict on the whole plate: pass or fail, from two tests. One is the spread of all Cq values; the other counts the wells that sit too far from the median. Beside it you read the mean, median, standard deviation and coefficient of variation, and the outlier wells by position.

A heat map draws the plate with the average of every row and every column beside it, so a warm edge or a drifting row shows before you read a number. Click a well to compare its Cq with its row, its column and the plate. Save the analysis and it stays with the experiment, with the limits it was judged on; its verdict, a Cq per well and the heat map go into the PDF and HTML report.

AH112Outer ring+0.3 cycles19% lower quantityAH112Outer ring +0.3 cycles
Fig. 1A 96-well test plate shaded by Cq: the second ring runs 0.1 cycles later than the center, the outer ring 0.3 and the corners 0.9. The four corners are the only wells more than 0.5 cycles from the plate median. Illustrative data.

Is the free plate homogeneity tool enough?

Often, yes. The free tool runs the same two tests on the same engine. The experiment is worth it when the test should leave a record, or the plate has 384 wells.

One 96-well plate of Cq values you already haveThe free tool gives the same verdict, from a paste or a file, with no account
A heat map you build in a spreadsheetDrawn for you, with row and column averages, the outliers marked and the same limits every time
A 384-well plateThe experiment only
The plate set up, the curves and a saved resultThe experiment only: the test plate, its run file and the verdict stay together

The limits that decide it are numbers you can see and change. The app’s defaults are lenient; for a well-performing instrument Jan Hellemans advises stricter ones in Functional qPCR instrument validation.

Spread of all Cq valuesPasses at an SD of 1 cycle or less by default; Jan advises below 0.2 to 0.25
A well that straysAn outlier when it sits more than 0.5 cycles from the median Cq
How many strays still pass2 on a 96-well plate, 5 on 384, by default
Missing wellsLeft out of the verdict, so count them yourself; Jan’s advice is none

For a check at installation or after a service visit, set the stricter limits yourself.

How it works in Clarida

01In the app: Design › Homogeneity setup

The test plate, in one form

Pick the sample and the assay from your library, or name new ones, then the mix and the cycling protocol. Choose the plate and how many wells to fill, and Clarida writes the layout. Once the runs are ready or a file is imported, the sample, assay and wells are locked.

  • On what: 96 or 384 wells, every well by default

Read the method: Workflow layouts

02In the app: Import › Import & annotate

The run, linked to the plate you set up

Drop in the file your cycler wrote and link it to the planned run. Every well then holds the test sample and assay, which lets the analysis read the plate by position.

On the Import page: Import & annotate

03In the app: Analyze › Raw data

The curves behind every Cq

When the file brings its curves, you see each well’s amplification and melt, and choose whose Cq to use. Raw data QC flags a curve that fits badly or sits close to the noise, so one failed reaction is easy to tell from a pattern.

In full on the Analyze page

04In the app: Analyze › Plate homogeneity

A verdict on the plate, and where it fails

Pick the target, and the run when there is more than one, then click Analyze. Change a limit and the result clears until you analyze again; save the one you want to keep.

  • Maximum SD: 1 cycle
  • Outlier threshold: 0.5 cycles from the median
  • Max outliers: 2 on 96 wells, 5 on 384

The free tool: The free tool: plate homogeneity

Why it is done this way

Optics and temperature control both vary across a qPCR block, and either can shift Cq by position. A temperature probe checks the block alone; a plate filled with one reaction tests the instrument the way your experiments use it. Jan Hellemans advises the test at installation, after maintenance, when results look off, and every quarter.

Use an assay you trust and a sample that gives a Cq around 25. Much earlier values are not representative of real experiments, and late ones carry the sampling noise of few template copies, which looks like a poor instrument but is not (Ståhlberg and Kubista, 2018).

The two tests answer different questions. The SD measures the overall spread; the outlier count, against the median, catches a few wells that drift while the rest agree, which one SD can hide. MIQE 2.0 asks you to report the instrument behind your results (Bustin et al., 2025), and a saved test shows how it performed when you checked.

What it doesn’t do

A test for one row, one column or the edgeShown by the row and column averages; the verdict rests on the whole plate
A trend across tests or instrumentsNot in Clarida: each saved test is judged on its own
A formal qualification (IQ/OQ), block temperatureNot in Clarida: a functional check you run in your own lab

Frequently asked questions

Does it show an edge effect?

Yes, on the heat map: each row and column has its average beside the plate, so a warm edge stands out. The verdict itself has no separate test for one row or one column.

Can you analyze a run you already did?

Not from the file alone: the test plate is set up first. Enter the sample, assay and wells you ran in Homogeneity setup, then import the file and link it to that run.

Can you test only part of a plate?

Yes. Set the number of wells in Homogeneity setup; the analysis reads only those that hold a Cq.

Does excluding a well change the result?

No. The homogeneity analysis reads every well with a Cq, whatever was excluded in Raw data, so the test always judges the instrument on the whole plate.

Updated

Test your cycler on one plate.

Pick a sample and an assay you trust, fill the plate and import the run. 4 experiments in your first month, then 2 a month. No card.

Other experiments

All kinds of experiment
  • Gene expression

    Compare expression between conditions: each target corrected for its own efficiency, normalized on several reference genes or the global mean, and tested, with a fold change and its interval.

    Start from
    Relative quantification preset
    From a run file
    Yes, analysis only
  • Reference genes

    Validate a panel of candidates in one experiment: design it, run it, rank stability by geNorm M and V and by NormFinder, and save the ranking for your next normalizations.

    Start from
    Reference gene finder preset
    From a run file
    Yes, analysis only
  • Present or absent calls

    One call for every sample and target, made only when all replicates agree, with positive, negative and internal controls checked beside it and the cut-offs kept with the saved result.

    Start from
    Detection preset
    From a run file
    Yes, analysis only
  • Genotyping

    A genotype table from KASP, TaqMan, DMAS or melt assays: each marker declared once, each call with its quality, and known controls that catch swapped alleles.

    Start from
    Genotyping preset
    From a run file
    Yes, analysis only
  • Multi-plate studies

    Several plates as one study: sessions on different days, every run file matched, and each correction shown with its standard error.

    Start from
    Relative quantification preset
    From a run file
    Yes, analysis only