FIGURE 2.
Structurally similar sequences are clustered together in the RESS. (A) Three synthetic structures designed for this analysis. (B) PCA of the structure sequences after projection to the RESS separates the sequences based on structure. (C) Distributions of the distances between pairs of related structure (“1-hp versus 1-hp,” “2-hp versus 2-hp,” and “3-hp versus 3-hp”), pairs of different structure (“Diff structs”), and pairs of random sequences (“Rand versus Rand”). Distance between pairs was calculated by Spearman distance (left panel) or sequence identity (right panel). Related structure pairs were closer, on average, than different or random pairs in the RESS.
