# 2026.7 · Data in, data out

> RDML in and out, an ELN export that carries the whole analysis, and the Reference gene finder and plate homogeneity check on your experiment's data.

Source: https://clarida.bio/changelog/2026-7
Updated: 2026-07-06

[Release notes](https://clarida.bio/changelog)

RDML in and out, an ELN export that carries the whole analysis, and the Reference gene finder and plate homogeneity check on your experiment's data.

By [Jan Hellemans](https://clarida.bio/changelog/2026-7#about-the-authors)

July 6, 2026 5 min read

Last month was about trusting your numbers. This month is about moving them. qPCR data lives in many tools, and until now getting it into and out of Clarida meant leaning on a variety of Excel tables. With 2026.7 we speak RDML, the community standard for qPCR data exchange, so you can bring in runs and whole experiments, including the raw amplification and melt curves, and hand a well-formed RDML file back to any other tool that reads it. The ELN export now carries the analysis as well as the design.

Two free tools now run inside the app. The Reference gene finder and plate homogeneity check now run on your experiment data in place, no copy-paste round trip. And your cyclers, cycling protocols and reaction mixes are now saved in the Library.

Bring it in. Take it out.
Jan

## New features

### Start an experiment from Cq values

Clarida has always let you jump straight into analysis without first building a design and executing a run, but the old entry screen did not make that obvious. Every new experiment looked as though you were expected to walk the whole workflow even when all you had was a table of Cq values. The reworked launchpad fixes that impression. It presents your starting points as equal, focused entry cards: "Build from scratch" or "Start from Cq values".

![The 'Start a qPCR experiment' launchpad with two equal entry cards side by side. 'Build from scratch' lights up the full Define, Design, Execute, Analyze, Report pipeline. 'Start from Cq values' grays out Define, Design, and Execute and highlights only Analyze and Report, making an analysis-only experiment an equal first-class choice](https://clarida.bio/changelog/media/2026-7-qPCR-launchpad.png)

### Save instruments, protocols and mixes

Recurring lab assets no longer have to be re-entered every experiment. Alongside the existing sample and assay libraries, Clarida now saves your **instruments**, **cycling protocols**, and **reaction mixes** in the Library. Register and rename your thermal cyclers, save thermal cycling programs and save them as new, and build reusable master-mix recipes, then pick any of them in-workflow when you set up a run.

![The Cycling Protocols storage library with search, filter, and sort controls and a New protocol action. 'My library' is empty with a prompt to save an experiment one-off or clone a template, above a row of Clarida-provided protocol cards (Standard qPCR 40-cycle, Fast 2-step 40-cycle, Standard qPCR with Melt Curve) each drawn as a temperature-versus-time thermal profile with usage counts like 'Used in 151 runs across 43 experiments'](https://clarida.bio/changelog/media/2026-7-stored-cycling-protocols.png)

Saved items are promoted into your library and reused across experiments, with usage counts and lifecycle controls so retiring an old cycler or archiving a protocol never disturbs the runs that already reference it.

### RDML, all the way

qPCR data rarely starts and ends in one tool. Clarida now reads and writes [RDML](https://pubmed.ncbi.nlm.nih.gov/19223324/), the open Real-time PCR Data Markup Language standard, so an experiment can move between your instrument software, other analysis packages, and Clarida without a lossy detour through spreadsheets. You can import a single run's data, or a whole experiment tree with the same fidelity as the qbase+ import, and export any experiment back out as a valid RDML file.

![The Report section of an experiment showing two export cards side by side: Export ELN ('Download a complete electronic lab notebook of this experiment', with a Download ELN action) and Export RDML ('Download as RDML, the open interchange format for qPCR data', with a Download .rdml action)](https://clarida.bio/changelog/media/2026-7-RDML-export.png)

This is also the release where Clarida stores real raw curves. Importing RDML now persists the per-well amplification and melt curves alongside the Cq values, and multiplex assays are mapped correctly to the right dye channel and target so complex runs come in with accurate assay and target information. RDML files that already carry an externally corrected Cq are honored as-is instead of being silently recomputed.

### An ELN export that includes the analysis

Your ELN export used to capture the design of an experiment as it was set up and executed. It now captures the whole thing. Exporting an experiment produces a self-contained, human-readable workbook that adds the full analysis lifecycle: per-well Cq measurements, derived qPCR datapoints, the processing settings that were actually applied (efficiency, normalization, inter-run calibration), quality-control verdicts, and an index of your saved analyses with the input data behind each figure.

![The Results sheet of an exported experiment workbook: a Saved Analyses index listing three stored analyses (an NRQ bar chart, a gene-gene correlation, and a Mann-Whitney U test) with their titles, results, and creation dates, above the input data behind the 'Palm: downregulated?' NRQ bar chart as a per-sample table of relative quantities and standard errors; a sheet tab strip at the bottom shows Overview, Execute, Data, Processing, QC, and Results](https://clarida.bio/changelog/media/2026-7-ELN-export.png)

Custom sample properties and previously dropped design fields, such as per-sample concentration, expected Cq, and negative-control thresholds, now flow through as well. The result is an offline record complete enough to serve as lab-notebook documentation or as supporting material for a publication, with empty sections pruned so the export stays readable for small experiments.

### geNorm and plate checks in experiments

Two of the free community tools now run directly inside the app. The Reference gene finder (geNorm) evaluates candidate reference genes for stability, and the plate homogeneity check surfaces spatial artifacts across your plate, both without exporting your data, reformatting and re-entering it in a separate tool. They live in the Statistics & apps area and read straight from your experiment.

![The Reference gene finder running inside the app on live experiment data: a grid of ten candidate reference genes (ACTB, B2M, GAPDH, HMBS, HPRT1, RPL13A, SDHA, TBP, UBC, YWHAZ) as columns and case/control samples as rows, each showing its Cq value, with per-column and per-row include/exclude toggles; below, a recommendation panel reads 'Recommended: Use 2 reference genes. Most stable genes: SDHA, HMBS'](https://clarida.bio/changelog/media/2026-7-geNorm-integrated.png)

## Quality of life

- **Improved fullscreen navigation** · Previous step added for back and forth navigation. The Analyze section gained a Back-to-overview navigation to support switching between the overview and the details of a selected analysis.
- **Steadier NRQ charts** · The NRQ bar chart keeps its toolbar, view, and footer stable as you step through targets, with a prev/next stepper for moving between them.
- **Clearer normalization empty state** · The Normalization section now shows clear actions when no reference genes were selected and auto mode fell back to None as the normalization method.
- **More reliable outlier median** · Plate homogeneity's even-length median now correctly averages the two middle values in its range statistics.
- **File-load errors are shown** · When an experiment file fails to load, the error now surfaces in the dialog instead of failing silently.
- **Deep links to workflow steps** · Workflow sections are now deep-linkable, so you can bookmark or share a link that opens a specific step directly.

## On the horizon

Now that RDML import brings in your raw amplification and melt curves, the next step is letting you see them. We are building an interactive Raw data section to navigate, slice, and visually inspect amplification and melt curves across a multi-run experiment, so you can judge data quality before you ever trust a Cq.

## Changes that may affect you

Some results change. What to re-check is below.

Added on October 30, 2026: changes we left out at release.

**Plate homogeneity's median was off for even-length sets.** Its outlier detection (IQR, modified Z, Hampel and MAD) took one middle value instead of averaging the two. Wells close to a threshold may now be flagged, or no longer flagged; re-run a plate homogeneity check you reported.

**RDML files with an externally corrected Cq keep it.** Clarida now uses those values as they are, with inter-run calibration set to External, instead of recalculating them. An experiment imported from such a file before this fix gives different numbers once imported again.

## About the author

- Jan Hellemans

  Co-founder, Clarida

  Leads the science and the methods at Clarida. Published the qBase method in 2007 and co-authored the MIQE guidelines.

  [Google Scholar](https://scholar.google.com/citations?user=L8EeDgwAAAAJ)

## Bring an RDML file in, with its raw curves.

Clarida imports a run or a whole experiment from RDML, and exports any experiment as an RDML file.

[Start free](https://app.clarida.bio/login) · [See the Import step](https://clarida.bio/qpcr/import)
