Overview
A publication-quality figure is the primary vehicle for communicating scientific results. Journals impose strict requirements on resolution, file format, color mode, dimensions, and font size. Beyond meeting technical specifications, effective figures tell a clear visual story, guiding the reader from the overall result to the critical detail without ambiguity. The process involves selecting the appropriate chart type, refining aesthetics, ensuring accessibility, and adhering to journal-specific guidelines. Software such as Adobe Illustrator, Inkscape, and R with ggplot2 are frequently used for final figure preparation.
Key Concepts
Resolution and format are foundational. Journals typically require 300 dpi for color figures and 600 dpi for grayscale or line art, saved as TIFF or EPS. Color considerations include using CMYK rather than RGB for print journals, maintaining sufficient contrast, and selecting colorblind-friendly palettes. Typography demands consistent font choice (typically Arial or Helvetica), legible axis labels (8–12 pt), and mathematical symbols in proper typefaces. Multi-panel figures require careful composition: aligned axes, shared legends, and panel labels (A, B, C) in a logical reading order.
Applications
Well-crafted figures elevate research across all biological disciplines. Structural biologists prepare high-resolution renderings of protein structures with clear secondary structure coloring. Microscopy techniques generate multi-channel images that must be presented without misleading adjustments to contrast or color balance. Flow cytometry data requires careful gating visualization and compensation controls in publication-ready dot plots and histograms.
Best Practices and Tools
Begin each figure by selecting the right tool for the data type. For statistical plots (bar charts, boxplots, scatter plots), R with ggplot2 or Python with seaborn/matplotlib provides the greatest control over every visual element. For multi-panel figure assembly and final touch-ups, use Adobe Illustrator or the free alternative Inkscape, vector-based editors ensure text remains sharp at any resolution. For color scheme selection, follow established guidelines: use ColorBrewer 2.0 (colorbrewer2.org) to generate perceptually uniform, colorblind-friendly palettes. The “viridis” and “magma” colormaps (built into matplotlib and seaborn) are excellent defaults for heatmaps due to their perceptual uniformity and grayscale reproducibility. For accessibility, avoid red-green combinations; use blue-orange or a single-hue sequential palette instead. When assembling a multi-panel figure, ensure all axes are aligned, font sizes are consistent (8-12 pt for labels, 6-8 pt for tick marks), and panel labels (A, B, C) are placed in the upper left corner of each subpanel in 12-14 pt bold. Export at 300 dpi (600 dpi for line art) in TIFF format with LZW compression. For example, a volcano plot of RNA-seq results should use: ggplot2 with geom_point(), alpha=0.3 to handle overplotting, dashed horizontal lines at p = 0.01 (adjusted), labeled points for the top 10 most significant genes using ggrepel for non-overlapping text, and a color gradient from gray (non-significant) through blue to red (highly significant and highly expressed), all on a white background with no gridlines.
