2026.11·From plate to report
All releases · November 2, 2026 · Jan Hellemans
Last month the plate editor gave you a free way to lay out a plate before the run. This month it moves into the qPCR app: Run layout is now edited directly on the plate, and every plate in Clarida, from pipetting to plate homogeneity, uses the same view.
At the other end of the experiment, the report builder is here. It turns an analyzed experiment into a report you can attach to an ELN entry, a manuscript or a lab meeting, and it checks the report before you download it: results that are out of date, a missing conclusion, a table nobody will read.
In between, your settings got more honest too. Every setting that differs from the experiment's settings set now says so, where you edit it. And when you choose reference genes, NormFinder now gives geNorm a second opinion.
Lay it out. Write it up.
Jan
New features
A report you can attach as it is
An analysis is only finished once you can show it to someone. The new Report builder in the Report phase starts a report from a template, shows it as a live preview while you put it together, and lets you download it as a PDF or HTML file in A4 or Letter. Clarida offers five templates and lets you create your own variants.
The report states what was actually done. Its methods name the cycling protocol and mix each run used, and its settings use the same names as the app. It draws on the figures and results you saved during the analysis, the plate and cycling protocol, and the raw-data curves you saved. Before you download, Checks lists what deserves a second look: results that are out of date, a missing headline or conclusion, a conclusion copied from generated text that has since changed, and tables too long to read. The checks never block a download.
Write the headline and conclusion once: they are saved on the experiment and appear in every report you make from it. The outline lets you add headings, move, rename and remove items. Print preview shows the real PDF with its page breaks. Each saved report keeps its versions, and a template you adapt can be saved as your own under Report templates in the Library, for the next experiment.
Don't like the look? Theme offers five: Standard, Briefing, Lab record, Airy and Preprint. Clarida's own templates start in Briefing.
Lay out your runs on the plate
The plate editor from 2026.10.1 now comes to Run layout in the qPCR app. In the Manual layout mode, switch the plate to Edit and lay out each run the way you would in the free plate editor: one well, a replicate group or a whole set per click, with undo and redo. Each run is a pill above the plate, where you can rename it, clear its plate or delete it.
Checks show the same findings as in the plate editor, such as wells with only a sample or only an assay, or assays without a negative control. Every change is saved as you make it, and the analysis waits until you are done: an "Analysis out of date" notice counts down and updates the results 30 seconds after your last edit, or right away with Update now. Runs that already hold data open in View, so nothing changes by accident.
For runs on QuantStudio and Bio-Rad CFX instruments, Instrument file in Start Run downloads the run as the plate setup file the instrument software imports, so the layout reaches the instrument without being typed again.
A second opinion on your reference genes
Choosing reference genes on the word of one method is a risk worth avoiding. NormFinder (Andersen et al., 2004) now runs next to geNorm, both in the free Reference Gene Finder and in the Reference gene finder section of the qPCR app: one analysis, with a geNorm and a NormFinder tab. geNorm ranks genes by how well they track each other; NormFinder compares their variation within and between groups of samples, so a gene that shifts with the treatment does not look stable just because the other genes shift too. Compare methods shows both ranks side by side and marks where the methods disagree. Clarida never merges them into one rank.
In the qPCR app, Group samples by picks the sample property that defines the groups, such as treatment or tissue, and the analysis uses your targets' PCR efficiencies. It is available to everyone who can open the experiment, like geNorm. In the free tool, the data grid now also takes sample groups, a PCR efficiency per gene, technical replicates (rows with the same sample name are averaged) and missing values.
Improvements
- See which settings you changed · Every setting that differs from the experiment's settings set now shows a marker. Click it to see who changed it and when, the set's value, and Reset to set value. Where wells or targets have their own values, the setting says how many. The settings card in Define counts the overrides and lists them, each with a jump to where it is set.
- Switch an experiment's settings set · Change settings set in Define moves an experiment to another set. You choose whether to keep your overrides or drop them, Clarida lists what each override becomes, and the results are recalculated.
- One plate view everywhere · Run layout, Fill plate, the RDML data section and Plate homogeneity now share the plate editor's plate, with the same colors and well marks. The free Plate homogeneity tool now accepts 384-well plates and well names such as A01 or a1.
- Instrument setup files · The plate editor's Export can now download a plate as a QuantStudio plate setup file or a Bio-Rad CFX plate file, ready to load into the instrument software.
- Getting started in the plate editor · A short guide now takes you through the first steps: samples, then assays, then the plate. A row of names copied from Excel is read as a list of names. If you click an export while signed out, it now continues by itself after you sign in.
- Two choices, one toggle · A setting with exactly two options is now a toggle in the side panel instead of a dropdown, so both options are visible at once.
- Emails that look like Clarida · Invitations, workspace requests, billing and connector emails now share one clear layout, with a plain-text version for every email.
Fixes
- Assigning a mix in one run changed other runs · Assigning or clearing a well's mix in Run layout also removed that well's assignment in every other run of the experiment, and could delete that run's mix plan. Only the edited run changes now.
- Excluding many measurements at once failed · Excluding or including a large selection, by hand or through the replicate QC sweep, failed on large plates. It now works whatever the selection size.
- 384-well labels in the RDML tool · The free RDML tool labeled every plate as if it had 12 columns, so on a 384-well file reaction 13 read B1 instead of A13. Wells are now labeled by the run's own format, in the table, the curve names and the hover.
- ELN export showed zero volumes · When an experiment's mix plan was missing, the ELN export wrote zero volumes, and groups using the default mix had no Mixes section. Both are now exported, and a 0% overage is no longer exported as 10%.
- Sample codes out of order on the plate · Samples added in one go, by a paste or from the plate editor, could get codes that did not follow their order: the fourth sample on the plate could read X6. Codes now follow the samples' positions, and an experiment opened from the plate editor keeps the editor's order. In an existing experiment whose order of positions differs from the order samples were added, the codes and their colors on the plate may change once; the samples and wells themselves do not.
- Your normalization choice stuck · Running the relative quantification could overwrite a normalization method you had just chosen by hand. Some experiments copied long ago were refused by the relative quantification altogether. Both now run as expected.
Changes that may affect you
One calculation changed: geNorm now handles missing values and PCR efficiencies, which can change reference gene stability on incomplete data. Everything else is about which settings reach the calculations: every analysis now reads its settings from the experiment's own record, and in four places that record and the analysis did not agree. All four were bugs.
geNorm handles missing values and PCR efficiencies. geNorm used to drop every sample with a missing Cq and assumed a PCR efficiency of 100% (E = 2) for every gene. It now excludes samples or genes one at a time, keeping as many measured values as possible, and corrects each gene's Cq values by its PCR efficiency: in the qPCR app the targets' efficiencies, in the free Reference Gene Finder the ones you enter. In the free tool, rows that share a sample name are now averaged as technical replicates. On complete data with E = 2 the M values are unchanged, and analyses you saved before keep their results.
Cq runs used default QC margins. Each Cq run checked the margin flags against Clarida's default thresholds instead of the ones set for the experiment, until you next edited a threshold. Cq runs now always use the experiment's thresholds. If you changed these thresholds, the margin flags can change at the next Cq run.
Inter-run calibration QC under the SD metric. With replicate QC set to the standard deviation, the inter-run calibration warning threshold still followed the hidden ΔCq threshold. Under SD it now uses the 0.5 Cq default. Experiments that use the SD metric and once stored a ΔCq threshold may get different inter-run calibration QC verdicts.
Genotyping calls are recalculated once. Changing the replicate QC metric or threshold did not update the replicate flags in genotype calls, which were still shown as current. Every genotyping experiment recalculates its calls once after this release.
Collect Material volumes come from the mix plan. The per-sample, per-assay and per-reagent volumes in Execute are now calculated together with the mix plan. Before, split groups counted only one part of the plate, reagents with the same name in different mixes were added together, and a fixed 10% overage was applied. Experiments still being planned recalculate their mixes once when you first open them after this release. The ELN export's Assay List gains a Volume column.
The Excel export's Processing and QC sheets now state the inter-run calibration threshold, normalization method, efficiency, inter-run calibration and scaling settings the analysis actually used, where those rows were blank or read "none" before.
