A molecular model can contain several chains, residues, interaction regions, or data values—and color is often the quickest way to tell them apart. It is also easy for color to create confusion: a smooth gradient may imply an ordered scale where none exists, while a set of unrelated colors can make a changing value difficult to follow.
A useful starting point is to ask what the colors need to communicate. SAMSON’s default palettes are grouped into discrete, sequential HCL, qualitative HCL, diverging HCL, and flexible diverging HCL types. That organization offers a practical way to choose a palette by the kind of information you are showing, rather than by which swatch looks most attractive.
First decide what the colors mean
- Separate categories? For distinct groups that do not have a natural order, start with a discrete or qualitative palette. Examples in the defaults include Okabe-Ito, Carto Safe, Set2, and Qualitative – Default. The goal is for neighboring categories to look like different labels—not like successive points on a scale.
- Show a low-to-high progression? For an ordered quantity, consider a sequential palette. Viridis, Blue, Heat, and Gray are among the available choices. A sequential scheme gives the viewer a visual path from one end of the range to the other.
- Compare values around a meaningful midpoint? A diverging palette can distinguish values on either side of a reference point. The default diverging HCL examples include Berlin and Blue-Red; the flexible diverging list includes palettes such as RdBu and TealRose. Use this approach only when the center of the scale matters to the interpretation.
These are not rules about which palette is scientifically correct for a particular dataset. They are a way to make the visual encoding match the structure of the information. If you are coloring separate components, avoid implying that one component is “more” than another. If the colors represent a measured or calculated range, make sure the palette reads as an ordered progression. If two sides of a comparison matter, make the midpoint easy to recognize.
Make the palette work in the scene
A palette that looks clear as a tiny strip may be harder to read across a dense molecular structure. Check it in the actual view, at the scale where people will inspect or share the model. Look for regions that blend together, colors that are difficult to distinguish, and gradients that compete with the molecular geometry. For categorical coloring, use only as many distinct colors as you need; unnecessary categories add visual noise.
For an ordered scale, try a few sequential options rather than choosing by name alone. Here is the default Viridis palette as one example:
For a diverging comparison, the two arms of a palette communicate which side of the midpoint a value falls on. SAMSON’s documentation notes that you can reverse the left and right arms in the color palette or color scheme dialogs. That can help when the direction of the mapping should be swapped, while keeping the same palette structure.
A quick check before sharing
- Write down what each color represents: a category, an ordered value, or a difference around a reference.
- Choose a palette family that fits that meaning, then compare a couple of the available palettes.
- Inspect the colored structure—not just the palette strip—and check that important regions remain distinguishable.
- If the mapping feels backward, try reversing the palette arms. If the defaults do not fit the message, SAMSON also supports creating custom HCL palettes.
This small check can prevent a common presentation problem: a model that looks polished but leaves viewers guessing what the colors mean. For more details and the complete default palette lists, see the SAMSON color palettes reference. You can also explore Colorizing and Visual presets.
Note: SAMSON and all SAMSON Extensions are free for non-commercial use. Get SAMSON.