[t0mdavid-m]

What it does

The Explore results page no longer wires the experiment selector, view bars and experiments by hand. It describes its data in a ViewerSpec, and the template's show_viewer draws the page.

The spec has two kinds of dataset:

• Runs: the top-down catalog, the Deconvolution, Native and Lysate views, the run's own default view, and render_grid. Each experiment keeps its own StateManager, so 📌 and ↺ still work.
• Comparisons: the comparison catalog, the Quantification, Identification and QC views of its kind, and the comparison grid.

The following works as before:

• Edit comparison, History, Add analysis and Export sit beside the selector.
• Views a run was not processed for are marked with a +.
• The page opens on a comparison just added, and on the newest run after a run finishes.

Single run | Comparison

A comparison is listed once. The Single run | Comparison switch beside the selector shows either the comparison as a whole or one of its runs, picked in a run selector, with the run views (Deconvolution, Native, Lysate). This is the template's switch, so OpenDDA and the template example behave the same.

Edit comparison as a dialog

Edit comparison opens the template's dialog, with one tab per step: Runs and conditions (which runs are compared, their condition and replicate, and the level: features, proteoforms or proteins), Filter, Missing values, Normalize and Test. Start over with the recommended settings resets them; Done computes the comparison again.

Label-free example

The label-free example (spike-in 1x vs 0.1x) now ships all four of its runs, so each one can be opened with Single run. To keep the example data small, every run is cut to the 45 s in which the proteoforms that differ most elute. The four label-free runs take 98 MB, compared with 73 MB for the one run shipped before. The comparison keeps five changed proteoforms, four higher and one lower in spike-in 1x.

The picked proteoform in every plot

• Volcano. It opens on a symmetric fold-change range, so the volcano is visible at first glance. A single picked proteoform, and the one picked by default, is ringed like a lasso selection. The selection buttons are in the plot toolbar, which also has lasso, measure, box zoom and home.
• Abundance per run draws every proteoform picked in the volcano (lasso, box or "All changed"), one line each, and names the proteoform (protein and range) or the feature (its mass) under its title.
• Feature intensity, CV and the other box plots mark the picked feature or proteoform on its box, and the axis range keeps the mark in view.
• PrSMs. Picking a PrSM in "PrSMs of the selected identification" updates the sequence view and spectrum; when the PrSM's run has no spectrum in the workspace, a line says so and names the PrSM shown instead.

Lysate view

The tables of the Lysate view are two to a row and taller, so their columns are readable.

Plots from the template's visualization library

The plots OpenDDA draws the same way now come from the template's src/viz/, fed from FLASHApp's comparison results and kept in FLASHApp's per-dataset cache:

• GO enrichment (go_panel.py): category selectors, the number of selected proteins on a line of its own with Selection | All proteins, and paginated Insight tables.
• Comparison steps (comparison_steps.py): where values are missing, the values imputation fills in, the boxes per sample before and after normalization, and the p-value histogram. src/render/postprocessing_plots.py is gone; src/render/series.py only adds FLASHApp's cache folder and info texts to the library's show_series.
• Volcano, PCA and clustered heatmap take their settings from volcano_settings, pca_settings / most_variable and clustered_settings.
• Group colours come from the library (group_colors), so a condition has the same colour in every panel here and in OpenDDA.

The abundance profile and the comparison table stay in FLASHApp, which draws them its own way.

Cleaner viewer

• Info icons. Every component has an ⓘ in its top left corner, with what it shows and how to use it. The texts are in src/render/teaching.py. They replace the grey captions under the plots, the default-layout caption, the "Comparing … of …" line and the GO info box.
• Tables show their title and buttons on one line.
• Bar charts have no zoom or drag, and their title sits close to the bars.
• Selection tools only where they select. Plots that send no multi-selection hide box and lasso.
• Aligned rows. The volcano plot ends level with "Abundance per run".

Depends on

• Four layers for code, a standard Viewer page apps describe as data, Single run | Comparison and Edit comparison streamlit-template#584, the standard Viewer page with the switch, the Edit comparison dialog and the visualization library. It is merged into this branch.
• Help tooltips, heatmap settings and text axes, aligned volcano plot t0mdavid-m/OpenMS-Insight#107 (among others, SeriesPlot box and bar modes). This branch uses its development build from dev-wheels/, which must become a PyPI pin before a release.

Screenshots (example workspace)

Lysate view	Volcano and abundance per run, named proteoform

Label-free comparison	Edit comparison

Single run of the comparison	Picked feature on the box plots (QC: label-free)

Abundance per run of all changed proteoforms	PrSM picked, sequence view and spectrum follow

Screenshots of the library's group colours (PCA, clustered heatmap) are still to come.

Tests

• pytest tests/ passes: 1479 passed.
• pylint (errors only) is clean on the changed files.

🤖 Generated with Claude Code

https://claude.ai/code/session_01FkEGen2HSCa8BFE1JW4jxN

[coderabbitai]

Important

• Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: ee476562-f9e0-4b4a-b182-371d29cec7af

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

• X
• Mastodon
• Reddit
• LinkedIn

[t0mdavid-m]

The Linux desktop e2e failure (ValueError: mode must be 'lines', 'markers' or 'lines+markers', got 'bars') comes from develop, not from this PR's change.

• Since Draw the comparison and GO plots with openms-insight #282, src/render/go_view.py on develop draws the GO chart with SeriesPlot(mode='bars').
• requirements.txt pins openms-insight==0.7.5, which is the newest release on PyPI, and that release does not accept bars.
• No desktop build has run on develop since Draw the comparison and GO plots with openms-insight #282. The last one was on 7bb80b1, before it.

The installed app fails as soon as the comparison GO view is drawn, so this check will stay red on develop too until openms-insight releases the bars mode and requirements.txt pins that release. This PR does not touch either file.

---

Generated by Claude Code

[t0mdavid-m]

The test-nginx (full, amd64) failure is not caused by this PR. The kind cluster could not pull redis:7.4.12-alpine because Docker Hub answered 429 Too Many Requests, so the redis pod never became ready. I'll rerun the failed job once the rest of the run has finished.

---

Generated by Claude Code