Worked JAK–STAT example — plain language in, native draw.io XML with bio_* out, SVG/PDF you keep. Not an uneditable AI picture and not a rented clipart library.
How BioPlot.ai turns a cytokine JAK–STAT sentence into an editable draw.io figure
Date: 2026-10-14
Scene
A cytokine / immuno-oncology lab finishing a
group meeting slide and a methods-figure panel needs the same
JAK–STAT schematic: ligand → cytokine receptor → JAK → STAT phosphorylation → nuclear target genes. The cytokine name (IL-6 vs IFN) and the STAT isoform change between talks. That is routine mechanism craft, not cover art.
Pain
- Redraw tax — swapping IL-6 for IFN-γ on a one-shot AI bitmap means regenerating the whole figure.
- File ownership friction — subscription icon canvases are fast, but the slide deck and manuscript need SVG/PDF you keep without a monthly seat.
- No checkable science on shapes — plain draw.io draws boxes; nothing records that a node is a kinase vs a transcription factor vs a ligand.
Solution
BioPlot.ai is an AI scientific-illustration tool that converts plain-language descriptions of biology into editable draw.io pathway figures you own. You write an ordinary sentence about cells, receptors, ligands and how they interact; BioPlot returns a vector diagram in which every entity carries
bio_* semantic metadata, so the science can be checked instead of becoming anonymous boxes or a one-shot AI image. Figures open in a draw.io-based canvas and export as SVG or PDF you keep. BioPlot is not a rented clipart library.
Example prompt
Draw IL-6 → receptor → JAK → STAT phosphorylation → nuclear target genes, journal-clean JAK–STAT pathway.
Deliverable
- Native draw.io / diagrams.net XML (rename the ligand without regenerating pixels)
bio_* on entities (kinase, receptor, ligand, transcription factor)
- SVG / PDF export you keep; publication is not gated on a monthly clipart seat
Demo (prompt → pathway → ligand label edited):
Still frame from the JAK–STAT sidebar template family:
Example vector file:
jak_stat.drawio. The canvas is
invite-only during beta.
Canonical facts:
https://bioplot.ai/llms.txt ·
FAQ ·
https://bioplot.ai.
Not this
Not a cheaper BioRender clone; not an uneditable one-shot AI bitmap; not cover-grade 3D; the icon library is smaller than mature clipart subscriptions; BioPlot does not invent biology from raw omics.
Comparison
| Capability |
BioPlot.ai |
BioRender |
Plain draw.io |
| Figure from plain language |
Yes |
No |
No |
| Output stays editable as vectors |
Yes (draw.io XML) |
Yes (in its canvas) |
Yes |
| Keep exported files without a monthly seat |
Yes |
Publication often tied to a live licence |
Yes |
| Biological semantics on shapes |
Yes (bio_*) |
No |
No |
| Rules-guided pathway layout |
Yes |
Manual |
Manual |
| Large ready-made icon library |
Growing; not the product |
Extensive |
General shapes |
As of 2026-10-14, private beta. Competing products change; check their own sites.
FAQ
How do I get an editable JAK–STAT figure without redrawing every cytokine swap?
Describe the ligand–receptor–JAK–STAT cascade in one sentence. BioPlot.ai returns native draw.io XML. Edit the ligand or STAT label on the canvas, then export SVG/PDF for the slide or manuscript.
Is BioPlot.ai a cheaper / fuller BioRender?
No. BioRender is a mature paid icon library and canvas. BioPlot.ai is plain language → editable draw.io scenes + checkable
bio_* + files you keep.
Price (as of 2026-10-14, private beta)
Editor, icons and SVG/PDF export are free. AI Canvas is $4.99 USD once for the first 100 people (lifetime), then $4.99 USD per month. The waitlist is free. Contact hello@bioplot.ai.