The `ribbon drawing' popularized by Richardson (1981) provides an elegant method to visualize the folding and secondary structure of proteins. Several authors have developed computer programs to create ribbon diagrams, notably Lesk and Hardman (1982), Pique (1982), and Burridge (in Doolittle, 1985).
We introduced (Carson and Bugg, 1986) a method to model a protein ribbon in much the manner of a ribbon used to wrap presents: composed of many roughly parallel smooth threads running the length of the ribbon. This conceptually simple algorithm fits cubic polynomial `B-spline' curves to the peptide planes. Only a modest amount of vector algebra is needed as most modern graphics devices have B-spline curves provided as a basic primitive. This algorithm is employed by commercial molecular modeling packages such as Sybyl (from Tripos) and Quanta (from Polygen). This may have influenced another ribbon method (Priestle, 1988), although no references to earlier work are given.
Our original paper employed the built-in functions of the Evans&Sutherland PS300 to draw the curves using a single color for each thread. Due to limitations of the E&S spline implementation, one cannot break the threads into segments for a residue-based coloring scheme. Carson (1987) described how the individual splines may be computed, how the ribbon may be sectioned into residues, how the method may be extended to nucleic acids, and how solid shaded ribbon models may be produced on raster devices. This style has also been widely copied. The paper additionally gives the definition of a `ribbon space curve'.