Create publication-ready scientific figures

Any laptop. Vector editing is comfortable with 8 GB RAM; no GPU required.

Component	Role	Price	Link
Matplotlib 3.11.1	scripted, reproducible scientific charts	Free / 免费open-source Python librarychecked 2026-08-02	https://github.com/matplotlib/matplotlib/releases/tag/v3.11.1
Inkscape 1.4.4	vector assembly, annotations and final SVG or PDF export	Free / 免费GPL open-source vector editorchecked 2026-08-02	https://inkscape.org/release/inkscape-1.4.4/

Matplotlib journal style

figure.figsize: 7.0, 4.5
figure.dpi: 150
savefig.dpi: 600
savefig.bbox: tight
font.family: sans-serif
font.sans-serif: Arial, Liberation Sans, DejaVu Sans
font.size: 9
axes.labelsize: 9
axes.titlesize: 10
legend.fontsize: 8
xtick.labelsize: 8
ytick.labelsize: 8
axes.linewidth: 0.8
lines.linewidth: 1.2
pdf.fonttype: 42
ps.fonttype: 42
svg.fonttype: none

Render and inspect figure outputs

uv venv --python 3.12 .venv
. .venv/bin/activate
uv pip install 'matplotlib==3.11.1' pandas
python make_figures.py
test -s figures/main.svg
test -s figures/main.pdf
rg -n '<text|font-family' figures/main.svg
sha256sum make_figures.py figure.mplstyle figures/main.svg figures/main.pdf > figures/SHA256SUMS

Figure integrity review

Review the figure against its plotting data and caption. Check axis variables, units, scales, truncation, uncertainty representation, sample sizes, legend mapping, color accessibility, panel labels and whether annotations could imply unsupported causality. List defects and required corrections. Do not judge visual polish before data integrity.

Known pitfalls

• Editing plotted data marks manually in Inkscape breaks reproducibility; restrict manual edits to layout and annotation.
• Confirm the journal's width, font and file-format rules before final export.
• RGB colors and transparency can change during print conversion.
• A 600 DPI setting does not improve vector elements and can inflate raster layers.

Browser only. Best for biology schematics and slide-ready scientific illustrations rather than primary statistical plotting.

Component	Role	Price	Link
BioRender Academic Individual Academic Individual plan current 2026-08-02	licensed scientific illustration templates and publication export	$35/mo$35 per month billed annually; official academic pricing also lists $39 when paid monthlychecked 2026-08-02	https://www.biorender.com/pricing
Matplotlib 3.11.1	generate data-driven panels before composition	Free / 免费open-source plotting librarychecked 2026-08-02	https://github.com/matplotlib/matplotlib/releases/tag/v3.11.1

Composite figure handoff

data_panels:
  source: scripted_matplotlib
  export: SVG
  manual_value_edits: forbidden
schematic_panels:
  source: biorender
  license_receipt: required
final_checks:
  - panel_labels_match_caption
  - fonts_embedded_or_outlined
  - minimum_text_size_met
  - colorblind_check_passed
  - publication_license_valid
archive:
  - plotting_code
  - source_data_hashes
  - editable_figure
  - final_pdf
  - license_evidence

Known pitfalls

• BioRender Free does not grant publication use according to the pricing page.
• The annual individual plan is a substantial commitment.
• Keep license evidence with the submitted figure.
• Do not redraw statistical data by eye inside an illustration tool.