Understand itTrace one replicate: baseline → align to pulse → response.
Input: three files
↓Process: baseline division, alignment
↓Output: aligned traces
Choose toolsPython for preparation, SVG for linked views.
Python csvHTML + SVGNo LLM
Build itPaste the brief into OpenCode and check the two baseline edge cases.
OpenCode · development brief
Build a pulse-response viewer: align replicate traces to pulse start, normalise to explicit baselines, flag zero and missing baselines, show linked views and export.▎
Compare meristem responses to hormone pulsesSource & credit
Compare meristem responses to hormone pulses: Trace one replicate: baseline → align to pulse → response.
Align to the pulse, compare to the baseline. Owner-supplied illustrative interface; source figures credited below.
Schlegel et al. 2021, eLife 10:e70934 — CC BY 4.0 (https://iiif.elifesciences.org/lax:70934%2Felife-70934-fig2-v2.tif/full/full/0/default.jpg)
Owner-supplied illustrative interface; source figures credited below.
Build itPaste the brief into OpenCode and check the bent-branch length test.
OpenCode · development brief
Build a root polyline annotator with scale calibration, parent-branch IDs, undo and keyboard editing; compute polyline lengths and insertion distances, flag disconnected branches and export a branch CSV.▎
Turn a root image into a branching mapSource & credit
Turn a root image into a branching map: Trace one branch: trace → confirm parent → measure.
Trace the axes, keep the parent links. Owner-supplied illustrative interface; source figures credited below.
NASA / JSC APEX-08 — public domain (https://commons.wikimedia.org/wiki/File:PRO_Imagery_Submittal_-_APEX-08_(jsc2021e036657).jpg)
Owner-supplied illustrative interface; source figures credited below.
Build itPaste the brief into OpenCode and run the rectangle test.
OpenCode · development brief
Build a napari viewer for two founder-cell contours, shared-wall endpoints and nuclei over time; validate calibration and IDs, compute area and nucleus-to-wall distance per frame and export.▎
Follow founder-cell shape before root initiationSource & credit
Follow founder-cell shape before root initiation: Trace one frame: validate → measure → timing.
Two cells, one wall, every frame measured. Owner-supplied illustrative interface; source figures credited below.
Wolny et al. 2020, eLife 9:e57613, Fig. 1 — CC BY 4.0 (https://iiif.elifesciences.org/lax:57613%2Felife-57613-fig1-v2.tif/full/full/0/default.jpg)
Owner-supplied illustrative interface; source figures credited below.
Understand itTrace one frame: pair poles → geometry → align events.
Input: poles + events
↓Process: pairing, length, orientation
↓Output: timeline + CSV
Choose toolsTrackMate optional, JavaScript/SVG.
TrackMate (optional)JavaScript + SVGNo LLM
Build itPaste the brief into OpenCode and confirm the unpaired frame.
OpenCode · development brief
Build a spindle timeline: require paired poles and calibration, compute separation and orientation, mark gaps, keep reviewed anaphase/cytokinesis times separate and export.▎
Align spindle motion with mitotic eventsSource & credit
Align spindle motion with mitotic events: Trace one frame: pair poles → geometry → align events.
Understand itTrace one cell: review → annotate → summarise with declared denominator.
Input: cells
↓Process: review states, polylines
↓Output: CSV with denominator
Choose toolsSVG annotation, JavaScript lengths.
SVGJavaScriptNo LLM
Build itPaste the brief into OpenCode and verify the denominator.
OpenCode · development brief
Build a cilia annotation app with yes/no/uncertain per cell, calibrated ciliary polylines, undo and keyboard editing; report counts with the reviewed denominator and export.▎
Review early cilia assembly cell by cellSource & credit
Review early cilia assembly cell by cell: Trace one cell: review → annotate → summarise with declared denominator.
Unreviewed is not 'no cilium'. Owner-supplied illustrative interface; source figures credited below.
Hall et al. 2023, eLife 12:e79299, Fig. 7A — CC BY 4.0 (https://iiif.elifesciences.org/lax:79299%2Felife-79299-fig7-v2.tif/full/full/0/default.jpg)
Owner-supplied illustrative interface; source figures credited below.
Understand itTrace one barcode: validate joins → cells → state composition per embryo.
Input: tables
↓Process: embryo-aware joins
↓Output: linked views + CSV
Choose toolsPython for joins, SVG/HTML for views.
Python csvHTML + SVGNo LLM
Build itPaste the brief into OpenCode and check the shared-barcode case.
OpenCode · development brief
Build a lineage-and-state viewer: join cells by embryo and barcode, treat barcodes as supplied labels, show collisions, compare state composition and export.▎
Compare cell fate with developmental historySource & credit
Compare cell fate with developmental history: Trace one barcode: validate joins → cells → state composition per embryo.
A shared barcode is a label, not a parent. Owner-supplied illustrative interface; source figures credited below.
Saunders et al. 2023, Nature 623:782, Fig. 1 — CC BY 4.0 (https://media.springernature.com/full/springer-static/image/art%3A10.1038%2Fs41586-023-06720-2/MediaObjects/41586_2023_6720_Fig1_HTML.png)
Owner-supplied illustrative interface; source figures credited below.
Choose toolsPython for calculation, HTML/SVG for review.
Python csvHTML + SVGNo LLM
Build itPaste the brief into OpenCode and check both edge cases.
OpenCode · development brief
Build a ratiometric review: background-correct channels, compute ratios only with a positive denominator, convert to pH only inside a user-supplied monotonic calibration and export.▎
Understand itTrace one readout: design → compatible contrasts.
Input: factorial data
↓Process: grouping, compatibility
↓Output: board + contrasts
Choose toolsPython grouping, HTML/SVG board.
Python csvHTML + SVGNo LLM
Build itPaste the brief into OpenCode and verify the unmeasured cell.
OpenCode · development brief
Build a diet × microbe comparison board: show the full design, distinguish unmeasured from zero, allow contrasts only among compatible measurements, show replicates and export.▎
Compare diet and gut-microbe effects togetherSource & credit
Compare diet and gut-microbe effects together: Trace one readout: design → compatible contrasts.
Understand itTrace one species: follow reactions → inspect evidence per step.
Input: records
↓Process: filter by species, follow routes
↓Output: pathway + route JSON
Choose toolsJavaScript lookup, SVG compartments.
JavaScript MapSVGNo LLM
Build itPaste the brief into OpenCode and check species separation.
OpenCode · development brief
Build a pathway explorer from reaction records: keep yeast and mammalian branches distinct, draw compartment-aware routes, label unknown steps, inspect evidence per reaction and export routes.▎
Trace O-mannosylation across cellular compartmentsSource & credit
Trace O-mannosylation across cellular compartments: Trace one species: follow reactions → inspect evidence per step.
Follow the sugar, keep the species apart. Owner-supplied illustrative interface; source figures credited below.
Chiapparino et al. 2020, eLife 9:e61189, Fig. 1 — CC BY 4.0 (https://iiif.elifesciences.org/lax/61189%2Felife-61189-fig1-v1.tif/full/full/0/default.jpg)
Owner-supplied illustrative interface; source figures credited below.
Understand itTrace one factor: which views it explains → its top loadings → how its scores line up with metadata (including batch).
Input: model export + metadata
↓Process: select a factor, join scores to metadata, rank loadings per view
↓Output: variance heatmap, loadings bars, metadata association panel
Choose toolsPython for validating the export, HTML/SVG for linked views; MOFA+ as the upstream model.
MOFA+ (or any factor model) — upstream; the app never fitsPython csv — validation of IDs across the four tablesHTML + SVG — heatmap, bars, linked selection
Build itPaste the brief into OpenCode and confirm a missing modality stays distinct from a zero loading.
OpenCode · development brief
Build a factor-review app from factors.csv, loadings.csv, variance.csv and samples.csv. Show variance explained per factor and view as a heatmap, let the user select a factor, list its top positive and negative loadings ▎
Inspect what a multi-omics factor representsSource & credit
Inspect what a multi-omics factor represents: Trace one factor: which views it explains → its top loadings → how its scores line up with metadata (including batch).
What does Factor 1 represent? Owner-supplied illustrative interface; source figures credited below.
Argelaguet et al. 2020, Genome Biol 21:111 — MOFA+ Fig. 2, CC BY 4.0 (https://media.springernature.com/full/springer-static/image/art%3A10.1186%2Fs13059-020-02015-1/MediaObjects/13059_2020_2015_Fig2_HTML.png)
COS Biological Data Science group (Velten) (https://www.cos.uni-heidelberg.de/en/research-groups/biological-data-science)
Owner-supplied illustrative interface; source figures credited below.
Understand itTrace one feature: declare scale → inspect regions per sample.
Input: spatial tables
↓Process: per-sample scaling
↓Output: side-by-side maps
Choose toolsJavaScript and SVG.
JavaScriptSVGNo LLM
Build itPaste the brief into OpenCode and confirm no silent overlay.
OpenCode · development brief
Build a spatial comparison viewer: side-by-side maps per sample with declared units, overlay only with explicit shared coordinates, missing values distinct from zero, region selection and export.▎
Compare spatial expression patterns across conditionsSource & credit
Compare spatial expression patterns across conditions: Trace one feature: declare scale → inspect regions per sample.
Two samples, two coordinate systems. Owner-supplied illustrative interface; source figures credited below.
Dries et al. 2021, Genome Biology 22:78, Fig. 6 — CC BY 4.0 (https://media.springernature.com/full/springer-static/image/art%3A10.1186%2Fs13059-021-02286-2/MediaObjects/13059_2021_2286_Fig6_HTML.png)
Owner-supplied illustrative interface; source figures credited below.
Understand itTrace one ray: boundaries → intersections → thickness.
Input: section + boundaries
↓Process: calibrated intersections
↓Output: overlay + CSV
Choose toolsFiji for annotations, SVG viewer.
Fiji/ImageJ (optional)SVG + JavaScriptNo LLM
Build itPaste the brief into OpenCode and run the concentric test.
OpenCode · development brief
Build a section annotator for retinal boundaries: place calibrated rays, report thickness per layer along each ray, flag missing or crossed layers and export.▎
Review retinal layers and growing eye shapeSource & credit
Review retinal layers and growing eye shape: Trace one ray: boundaries → intersections → thickness.
Thickness along a ray, never a guess. Owner-supplied illustrative interface; source figures credited below.
Lust & Wittbrodt 2018, eLife 7:e32319, Fig. 6A — CC BY 4.0 (https://iiif.elifesciences.org/lax:32319%2Felife-32319-fig6-v2.tif/full/full/0/default.jpg)
Owner-supplied illustrative interface; source figures credited below.
.jpg){kind=link}
