Fig. 1.

This is a logo of a prediction of the str2 alphabet for a casp7 protein sequence that had no close PDB homologs, but which was predicted to consist almost entirely of anti-parallel β-strands and one α-helix. See text for explanation of heights of letters. The bars below the sequence positions represent segments of a single contiguous predicted letter of the full str2 alphabet. When the structure was solved, it turned out that this predicted secondary structure was very accurate.