A molecular structure can contain the information you need and still be difficult to read. A common reason is color: one palette is used for unrelated categories, a gradient is used where there is no natural order, or a meaningful difference disappears in a busy view. Choosing a palette by the kind of information you want to show can make a structure easier to interpret—and easier for someone else to understand at a glance.
SAMSON includes several families of default palettes. Instead of beginning with “Which colors look best?”, begin with “What does color represent here?” That question helps narrow the choices.
For separate categories: use a discrete palette
Use a discrete palette when colors distinguish separate groups rather than positions along a scale. For example, you might want visually distinct colors for different chains, components, or selected groups. The discrete collection includes palettes such as Accent, Okabe-Ito, Set1, and tab10.
The Okabe-Ito palette is one of SAMSON’s default discrete options.
A useful check is to look at the model and ask whether each color can be matched to one category without needing to infer a gradient. If several groups look too similar, try another palette or reduce the number of categories shown at once. Avoid assigning a different color to every small structural element unless those distinctions help answer a question; too many competing colors can make the overall structure harder to parse.
For ordered values: use a sequential palette
When color represents values that progress from lower to higher—such as a measurement or a mapped property—a sequential palette is usually the more natural fit. Its progression gives the viewer a visual sense of order. SAMSON offers options including Viridis, Plasma, Blue, Gray, and Green-Yellow.
Viridis is one of the available sequential HCL palettes.
Before settling on a gradient, check that its direction agrees with the meaning of the data: which end represents the smaller value, and which end represents the larger one? Make that relationship clear in a legend or caption when your workflow includes one. A palette name alone does not explain what the values mean.
For comparison around a reference: use a diverging palette
If the important story is how values differ on either side of a meaningful reference point, consider a diverging palette. This is often useful for showing deviations around a baseline, where the midpoint matters as much as the two extremes. SAMSON’s diverging HCL options include Berlin and Blue-Red, while its flexible diverging collection includes choices such as RdBu, PiYG, and Spectral.
Blue-Red is one of the default diverging HCL palettes.
Use this family only when the data really has a meaningful center. If your values simply increase from low to high, a sequential palette avoids implying a special midpoint that is not part of the analysis.
A quick selection routine
- Name the data. Is color encoding categories, an ordered quantity, or differences around a reference?
- Choose the matching family. Start with discrete, sequential, or diverging palettes accordingly.
- Inspect the structure at its usual viewing scale. A palette that looks clear in a swatch may be harder to distinguish across a dense molecular view.
- Check the meaning, not just the appearance. Confirm that the colors and their order support the interpretation you intend.
- Keep comparisons consistent. When comparing views, avoid changing palette meaning between them without making the change explicit.
The HCL palettes use the Hue-Chroma-Luminance color space. SAMSON also lets you reverse the left and right arms of palettes in the color palette or color scheme dialogs, and create custom HCL palettes when the defaults do not suit your task. For details on applying palettes in a colorizing workflow, see Colorizing; the related Visual presets page covers another part of preparing a clear view.
For the complete list of default palettes, see the SAMSON color palettes reference.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.