Workflow layouts

How to choose which sections a Clarida experiment shows, under which heading and in what order, for one experiment or for every experiment started from a preset, and how layouts work: what a layout holds, how it reaches an experiment, what the layout editor allows and why, and how layouts are locked, duplicated and updated in the Library.

An experiment in Clarida is a sequence of sections: Define, Design, Execute, Analyze and Report, each holding the steps of that stage. Most experiments use only part of it. A lab that imports finished runs has no use for Design; a genotyping study needs Marker Setup and Genotype calling and has no use for normalization.

A workflow layout decides which sections an experiment shows, under which heading and in what order. It decides what you see, never what is computed: hiding a section changes no Cq value, no normalization and no genotype, and showing it again brings back everything it held.

This page has two parts. How to covers changing the layout of one experiment and building your own. How it works explains what a layout holds and the rules behind the editor; read it when a step raises a question. The running example is a lab that types SNPs by TaqMan allelic discrimination and wants its own version of Clarida's Genotyping layout.

1. Change the layout of one experiment

Open Define. Under Active settings and layout, the layout card shows the experiment's layout, with its name and where it came from. Its three icons, from left to right:

Maximize

• View workflow layout shows the layout's description and an outline of its sections.
• Switch workflow layout applies another layout from the Library. Use it when the experiment was started from the wrong preset, or has turned into a different kind of experiment.
• Customise workflow layout… opens the layout editor on this experiment's own layout. Use for this experiment changes this experiment only; Use and save to Library also saves the result as a layout of your own. A customised layout reads Custom layout, and offers Save layout to Library… for later.

None of these asks for confirmation, because none of them loses anything: a section the new layout hides keeps its data, saved analyses and settings, and switching back brings it back as it was. Each change offers Undo.

See How a layout reaches an experiment.

2. Build your own layout

In the sidebar, open Setup → Workflow layouts, select the layout closest to what you need (for our lab, Clarida's Genotyping) and choose Duplicate layout. The editor opens on Genotyping (copy), a table with one row per section. Rename it, then work through the rows:

• Leave out what you never use. Set In this layout to Declined. Our lab declines Melt, because TaqMan allelic discrimination has no melt step, and Execute, because its bench work is recorded elsewhere.
• Gather sections under a heading of your own. Include a custom group (Group A), name it, and give it a parent. Our lab names it Genotypes and places it under Analyze.
• Move sections. Change a row's Parent, and use the up and down arrows to set its order. A moved row arrives last among its new siblings. Our lab moves the Genotyping report application and then Genotype calling into Genotypes, moves Genotype calling up one place so it comes first, and moves the group up to where Genotype calling used to stand.
• Check the result. Included only lists just the rows the workflow will show, in the order it will show them.

If a change breaks a rule, the problem appears under its row with the fix in the same sentence, and Save stays disabled until it is fixed.

Maximize

See The editor table, Included, declined and not yet decided and What can move, and where.

3. Start experiments from it

A layout reaches new experiments through a preset. Point your genotyping preset at the new layout (see Presets), and every experiment started from that preset opens with this structure. The first one locks the layout's structure.

See Locked once used.

4. Take in what is new

When Clarida publishes a newer version of the layout yours was copied from, or adds sections your layout has never mentioned, its details in Setup → Workflow layouts show Update available, the number of new sections, or the number of updates when there are both. That opens What is new for this layout, where you take each change or leave it. Nothing is applied for you, and experiments already started are never affected.

See Inheriting from a Clarida layout and Sections Clarida adds later.

What a layout holds

A layout lists sections, and for each one records three things:

• Whether it is shown. A section the layout includes appears in the workspace, in the footer's step pills, and in Next and Previous. A section it leaves out appears in none of them.
• Where it sits. Its parent: the phase, section or group it appears under.
• Its order among the other sections under the same parent.

A layout holds nothing else: no data, no settings, no samples. That is why one layout can serve many experiments, and why switching layouts on a running experiment is safe.

Layouts, presets and settings sets

Three Library items decide how a new experiment starts, and each answers a different question:

• A workflow layout decides which sections show, and where.
• An experiment settings set decides how those sections compute: the analysis, QC and design settings they use.
• A preset brings them together. It references one layout and one settings set, plus any samples, assays, mixes or protocols the experiment should start with.

Layout and settings are independent. A section can be hidden while its settings still apply to the calculation. When a result looks wrong, look at the settings set; when a screen is missing, look at the layout.

How a layout reaches an experiment

Every experiment starts from a preset, and every preset carries exactly one layout. Starting an experiment copies that layout into the experiment and records which Library layout the copy came from. A blank start uses Clarida's default preset, whose layout is All sections.

From that moment the experiment reads its own copy, so later changes in the Library never rearrange an experiment that is already running. Switch workflow layout replaces the copy with another layout as it stands in the Library today. Use for this experiment gives the experiment a layout of its own: Define then shows Custom layout, marked This experiment only, and nothing is added to the Library.

If the section you were looking at is hidden by a new layout, the workspace moves you to the nearest section that is still shown and says where.

Clarida's layouts

Clarida publishes a set of layouts that every workspace can use directly, and from which you can duplicate your own:

A preset that references one of these directly gets Clarida's improvements: when Clarida publishes a new version, the next experiment started from that preset gets it. To change one of them, duplicate it.

The editor table

The layout editor has one row per section, including the sections the layout does not show, so you can see what you are leaving out. Its columns:

• Section: the name, indented under its parent, with a badge where the kind of row is not obvious.
• In this layout: whether the section is shown.
• Parent: what it sits under, for a section that can move.
• Order: its position among its siblings, such as 3 of 6, with arrows to move it up or down.

Rows are drawn in the order the workspace will draw them, so a move visibly moves the row. Everything lists every row; Included only lists only what the workflow will show.

Maximize

Phases, sections and steps

The five phases (Define, Design, Execute, Analyze and Report) are the top level. A phase can be reordered or left out, but never moved inside another section. Define cannot be left out: it is where an experiment's layout and settings are chosen, and a new experiment opens on it wherever the layout places it.

Below the phases are sections, and below some sections their steps: Cq processing holds Exclusion, IRC, Replicates, Efficiency and Normalization; Raw data holds Amplification, Melt, Raw data QC and Stored views. Sections differ in what they can hold:

• Design, Analyze, Cq processing and Raw data draw whatever the layout places in them, in the layout's order.
• Execute and Report draw only their own steps, in a fixed order.
• Exploration and Statistics & apps are app pickers: each shows one of its own analyses at a time.
• Every other section holds nothing.

Badges mark rows whose kind is easy to misread: App picker on Exploration and Statistics & apps, and Only with raw curves (or Only with inter-run calibrators) on a section the workspace shows only to experiments that have those data. Such a row also has an Always show checkbox, for a layout that wants the section shown even when it has no data.

Custom groups

A custom group is a heading a layout names, gathering sections from any phase. It has no content, help or settings of its own. It shows the progress of what it holds, and disappears from the workspace when nothing in it is shown.

Clarida provides three, listed at the end of the editor as Group A, Group B and Group C with the Custom group badge. Including one places it at the top level, right after Define. You then name it, choose an icon for its header, and give it a parent: the top level, Design or Analyze. A group cannot sit inside another group.

Applications and app pickers

The specialist analyses (the statistical tests, the exploration charts, Reference gene finder, Plate homogeneity, Detection and Genotyping report) are applications. Each lives in an app picker by default and carries the Application badge. Its In this layout choice has three values:

• Shown with its host: listed in its picker, among the others.
• Placed on its own: it leaves the picker and becomes a section in its own right, which you can move like any other. It first lands right after its picker.
• Declined: not offered in the picker at all.

An application placed on its own keeps everything: its saved analyses are listed in its own section instead of in the picker. Clarida's Genotyping layout places the Genotyping report on its own under Analyze.

Alternates

An alternate is a second form over steps that already exist, for experiments that need them in a simpler shape. Homogeneity setup is the one Clarida ships: one form that sets up a plate homogeneity run (one sample and one assay in every well) and writes the same records Sample Prep, Assay Prep, Mix, Run Layout and Cycling Protocol would write. It carries the Alternate badge.

No layout shows an alternate unless it includes it. Included, it goes last under Design and can be moved like a Design step. Clarida's Plate homogeneity check layout includes Homogeneity setup and declines the five steps it replaces.

Included, declined and not yet decided

What the In this layout column offers depends on the row:

• Included shows the section. Declined hides it on purpose and for good: Clarida never suggests a declined section again. Use it for "we never do this here".
• New — not yet decided cannot be chosen. It belongs to a section Clarida added after the layout was made. Until you include or decline it, the workspace does not show it, and the Library keeps offering it.
• Not used is the resting state of a custom group or an alternate. They exist for the layouts that ask for them, so a layout that does not is not asked about them.
• Always shown: the workspace always shows this section, and a layout cannot hide it yet. Define and the four Execute steps read this way.
• Controlled by Clarida: the section is switched off for your workspace by Clarida, whatever the layout says. This applies to sections that are not yet available to everyone.
• Left out with …: the section sits under a phase or section the layout does not show, and goes with it. Declining Execute shows its four steps as Left out with Execute; including Execute again, before saving, puts them back as they were.

Sections Clarida adds later

When Clarida adds a section to the product, every existing layout has, by definition, never mentioned it. Layouts show only what they include, so the new section stays hidden in all of them until someone decides.

This is why Declined and New — not yet decided are different states. The layout's Library page counts the undecided sections and offers each one, with Add selected and Decline selected. A declined section is never offered again. Custom groups, alternates and applications are never offered: they are optional by design, not new.

What can move, and where

The Parent control appears only on a row that can move; a section that cannot move shows its fixed parent as text.

• What moves: the Design steps (including Homogeneity setup), and in Analyze: Import & annotate, Raw data with Amplification, Melt and Raw data QC, Cq processing with its five steps, Genotype calling, Sample/Assay QC, Exploration, Statistics & apps, and every application placed on its own. A section that holds steps takes them with it.
• What stays: the phases, which are reordered instead; Execute's four steps; and Stored views, which belongs to Raw data.
• Where it can go: the Top level, and every section that draws what is placed in it (Design, Analyze, Cq processing, Raw data and any included custom group), as long as the layout includes it and the section would still fit within three levels there.

So the list never offers Define, Execute or Report, which hold no sections from elsewhere, nor Exploration or Statistics & apps, which show only their own analyses. A section can leave its phase: Sample Prep can sit in a custom group at the top level. Its saved state, help and settings follow it, and its progress counts where it now sits.

Order. The up and down arrows move a row one place among its siblings. A row given a new parent arrives last among its new siblings. Execute and Report draw their steps in a fixed order, so their rows have no order control.

Three levels deep

A layout is at most three levels deep: a phase or top-level item (level 0), what sits in it (level 1), and what sits in that (level 2).

The rule counts the steps a section brings with it. Cq processing carries its five steps one level below it, so it can sit at level 1 but never at level 2. A custom group at the top level holds sections at level 1; a group under Analyze holds them at level 2, so a section with steps cannot go in it. Our lab's Genotypes group sits under Analyze and holds two sections without steps, which fits.

Rules the editor checks

The editor checks the whole layout as you edit. Most problems come from three rules:

• Steps that depend on each other keep their order. Marker Setup needs Assay Prep, because markers are derived from the assays' targets and dyes. Mix comes after Sample Prep and Assay Prep, because the mix is sized for samples × assays × replicates. Run Layout comes after Sample Prep, Assay Prep and Mix. The rule reads the workflow from top to bottom, across phases and groups, so moving Marker Setup into a group placed above Design is refused.
• A phase emptied by moves must be declined. If every section of Design now sits elsewhere, Design would show nothing and leave an empty stop in Next and Previous.
• A section's parent must be shown. A section placed under something the layout does not include is refused.

Two smaller ones: a custom group needs a name, and Define must be included.

Locked once used

Setup → Workflow layouts lists your own layouts, marked Mine, beside those shared with your workspace, such as Clarida's. A layout is locked once an experiment has been started from it. Its name and description stay editable; its structure does not. The edit panel says so: Started 3 experiments here — its structure is locked so those experiments stay traceable to it.

The lock keeps one meaning for "this experiment used this layout". To change a used layout, choose Duplicate layout: the copy is new, unused and editable, and presets that should use it are then pointed at it. A preset that has itself started experiments is locked in turn; duplicate the preset (see Presets).

A layout that nothing in your workspace uses, no preset and no experiment, can be deleted. One in use can be archived: it disappears from the Library and from every layout chooser, while the presets and experiments that use it keep working.

Inheriting from a Clarida layout

A duplicate is your own copy, and it never changes by itself. Its details say so: Derived from Clarida · Genotyping (version 2026-10-05). Your own copy — it does not update. It remembers where it came from, which is what lets Clarida tell it about a newer version of its source.

When something new concerns it, its details show Update available, the number of new sections, or the number of updates when there are both, which opens What is new for this layout. A newer version of the source is offered according to the lock:

• Replace with the newer version… on an unused copy. It takes the source's new structure and discards your own changes, so it asks first.
• Duplicate the newer version on a used copy, whose structure is locked. You get a fresh copy of the new version; your old copy stays as it was.
• Dismiss keeps your layout as it is and drops the notice until the next version.

New sections are offered with Add selected and Decline selected. On a used copy the list still shows, for information; duplicate the layout to act on it.

Publishing

Publishing is how a vendor, such as a kit maker, makes a layout available to other workspaces. Clarida's own layouts are published the same way. In a vendor workspace each layout has an audience (Private, Partners or Public), and a published layout behaves differently from a lab's own:

• It never locks. Saving a change to its structure publishes a new version instead of changing the layout in place.
• Experiments already started keep their copy. Presets that reference the layout start new experiments from the new version.
• Copies made from it in other workspaces are told Update available.

A lab workspace cannot publish: its layouts lock once used, as above.