Overview
A good scientific figure makes one point clearly, shows the data honestly and meets the technical requirements of the place it will appear. Many figures fail on one of these: a bar chart hiding the distribution, a rainbow color map that distorts gradients and excludes readers with color vision deficiency, text too small after reduction to column width, or a raster image where the journal expects vector artwork.
We support three kinds of work. Data figures are built in code (Python with Matplotlib or seaborn, R with ggplot2, MATLAB) from your data, so they can be regenerated when results change. Schematics and conceptual diagrams, such as experimental setups, workflows, mechanisms and study designs, are drawn as vector graphics in Adobe Illustrator, Inkscape or TikZ. Figure assembly and review covers panel layout, labeling, fonts, color and export for a specific journal or thesis.
Image integrity is non-negotiable. We never alter micrographs, gels, blots or other experimental images beyond adjustments applied uniformly to the whole image and disclosed as required, and we do not generate images that could be mistaken for real data.
Who it’s for and the problems it solves
This is for you if you are…
- PhD students and postdocs preparing figures for manuscripts and thesis chapters
- Researchers responding to reviewer or editor comments about figure quality or format
- Master's students who want their thesis figures to look consistent and professional
- Labs that want a shared figure style and reproducible plotting scripts
- Authors who need a graphical abstract, conceptual framework or methods schematic
Common problems we help with
- The journal returned your figures for low resolution, wrong format or wrong dimensions
- Figures made in different programs have inconsistent fonts, sizes and colors
- A reviewer said your color scheme is not accessible to colorblind readers
- Your bar charts with error bars hide the spread of the data
- You have a hand-drawn sketch of an apparatus or pathway and need a clean diagram
What’s included — and what isn’t
Included
- Chart selection advice: when to use scatter, line, box, violin, raincloud, heat map, forest plot or small multiples, and when bar charts mislead
- Reproducible data plots in Python (Matplotlib, seaborn), R (ggplot2, patchwork) or MATLAB, with scripts you keep
- Schematics and diagrams: experimental setups, workflows, study flow diagrams, mechanisms, conceptual frameworks, drawn in Illustrator, Inkscape or TikZ
- Multi-panel figure assembly with consistent panel labels, alignment, fonts and sizes
- Color: colorblind-safe palettes (for example Okabe–Ito and viridis family), perceptually uniform color maps, grayscale checks
- Technical specifications: final dimensions for single or double column, minimum font sizes, line weights, resolution in dpi at final size, RGB or CMYK
- File formats: vector (PDF, EPS, SVG) for plots and line art, TIFF or high-quality PNG for images, and embedding fonts
- Captions and legends: what they must state (sample sizes, error bar definitions, statistical tests, scale bars)
- Image integrity review: checking that processing of micrographs, gels and blots follows journal policies
- Graphical abstracts sized to the publisher's requirements
Not included
- Altering, enhancing selectively, splicing without disclosure or otherwise manipulating experimental images
- Creating data, images or results that do not come from your research, including AI-generated images presented as data
- Producing assessed figures for student coursework where the figure itself is being graded
What you receive
- Final figure files
- Figures exported at the required size and format, typically vector PDF or EPS for plots and diagrams and TIFF for image panels, with fonts embedded.
- Editable sources
- Plotting scripts, Illustrator, Inkscape or SVG files, or TikZ code, so you can update figures later.
- Style sheet
- A short record of fonts, sizes, line weights and palette, and a Matplotlib style file or ggplot2 theme where useful, for consistency across the thesis or paper.
- Figure review notes
- For review-only work, comments on clarity, chart choice, color, labeling, caption completeness and journal compliance.
How it works
Brief
You send the data or current figures, the message each figure should convey and the target journal or thesis guidelines.
Plan
We suggest chart types and a layout for each figure, and agree a style before producing anything.
Build
Data plots are coded from your data; schematics are drawn from your sketches and descriptions.
Review together
You check scientific accuracy and we revise. Only you can confirm that a diagram represents your method correctly.
Export and check
Figures are exported to specification and checked at final size for legibility and in grayscale and colorblind simulations.
Design principles and technical standards
- Show the data
- Plot individual points or distributions alongside summaries when sample sizes allow, so readers can judge variability themselves.
- Colorblind-safe color
- Palettes such as Okabe–Ito for categories and viridis or cividis for continuous data; never color alone to carry meaning, and no red–green contrasts as the only distinction.
- Vector first
- Plots and line art saved as vector graphics scale without blur; raster images are used only for photographs and micrographs, at sufficient resolution at final printed size.
- Resolution at final size
- Resolution requirements refer to the final printed dimensions. Many journals ask for around 300 dpi for photographs and higher for line art in raster form; check your journal's current artwork guide.
- Typography
- One sans-serif font across figures, consistent sizes, and text that remains legible after reduction; text kept as text rather than outlines where journals require it.
- Image integrity
- Brightness and contrast changes applied to the whole image and to controls equally, no cloning or healing tools, non-adjacent lanes marked, and original files kept.
Image integrity rules we follow
Journals such as those in the Nature Portfolio state that adjustments like brightness and contrast are acceptable only when applied to the whole image and equally to controls, that cloning and healing tools are not acceptable, that rearranged non-adjacent gel lanes must be marked with visible boundaries, and that unprocessed originals may be requested. Many publishers also restrict AI-generated images. We apply the strictest relevant policy, keep a record of any processing and flag anything in existing figures that could raise questions during review.
Getting technical specifications right the first time
Returned artwork delays publication. Before designing, we check the target's column widths, maximum height, font size range, accepted formats and color mode. Figures are then built at their final size, so fonts and line weights are correct without rescaling.
- Design at final width (for example single column or full page width), not larger
- Check text size at actual size on the exported file
- Use RGB for online-first publication unless the journal asks for CMYK
- Name files as the journal requests, for example Fig1.pdf
Consistent figures across a thesis
Thesis figures often come from years of different software and habits. A shared style file and palette, applied through scripts, makes chapters look like one document and makes late changes easy. We can set this up for Matplotlib or ggplot2 and restyle existing figures from their data, while formatting of tables and captions can be handled with our tables and figures formatting service.
What the work can look like
Illustrative example — fictional data
Sample figure review for a manuscript revision
A cell biology postdoc received a reviewer comment that Figure 3 was hard to interpret. The comments below are invented to show the format of a review.
Chart choice
Panel B shows mean ± SEM bars for four conditions with n = 6 per group. Replace with a dot plot showing each replicate plus mean and 95% confidence interval, and state in the legend whether points are biological or technical replicates.
Color and legibility
Treatment groups use red and green only, which many readers with color vision deficiency cannot distinguish. Switch to Okabe–Ito orange and blue with different marker shapes. Axis labels are 5 pt at final size; increase to at least 7 pt, following the journal's range.
Integrity check
Panel D splices two western blot lanes from different positions without a boundary. Add a visible dividing line and state the splice in the legend, and make sure the uncropped original is available for the supplementary information.
Illustrative example with invented details. Always follow your target journal's current artwork and image integrity policies.
What you’ll need to provide
- Raw or processed data in a tidy format (CSV, Excel, R or Python objects) and the analysis that produced it
- Current figures, sketches or reference images for schematics
- The key message of each figure and draft captions
- Target journal's artwork guidelines or your thesis formatting rules
- Original, unprocessed image files for any microscopy, gel or blot panels
You don’t need everything ready to ask for a quote — tell us what you have and we’ll say what’s missing.
What affects price and turnaround
Price depends on
- Number of figures and panels
- Data plots from clean data versus schematics drawn from sketches
- Complexity of diagrams, such as detailed apparatus or mechanism illustrations
- Whether reusable scripts, style files and editable sources are required
- Number of revision rounds and journal-specific formatting
Turnaround depends on
- Condition of the data and completeness of the brief
- Time needed for you to check scientific accuracy of diagrams
- Journal resubmission deadlines
Every project is quoted individually. How we price our work.
Questions people ask before starting
Can you make my western blot or microscopy images look better?
Only within image integrity rules: uniform brightness and contrast adjustments applied to the whole image and equally to controls, cropping that does not remove relevant information, and clear marking of splices, all disclosed as required. We will not remove bands, clone areas, enhance selectively or combine images in ways that change what they show.
What resolution does my figure need?
It depends on the journal and the content type. Resolution applies at the final printed size; photographs are commonly required at around 300 dpi, and raster line art at higher resolution, while vector formats avoid the issue for plots and diagrams. We check your journal's current artwork instructions and export accordingly.
Can you use AI to create my graphical abstract?
We do not use generative AI to create images that could be mistaken for data or real scenes. Many publishers restrict AI-generated imagery, so we create schematics and graphical abstracts with vector drawing tools and your input.
Will I be able to edit the figures later?
Yes. You receive plotting scripts and editable vector files, so you can update data, change labels or adapt figures for a presentation or poster.
I am a student. Can you make figures for my assignment?
If the figure is part of what is being graded, no, but we can review figures you create and teach you the tools. For theses, check your institution's rules on outside help with figures and acknowledge it if required.