Transfer of information requires alignments, preferably structure alignments but
as those normally are unobtainable, we will have to do with sequence
alignment. To make optimal alignments we need to understand the characteristics
of the amino acids. And here the secondary structure kicks in because these
characteristics are not independent of the secondary structure. Pro is good
as the first residue of a helix. Asn, Ser, and Asp can form H-bonds with the
own backbone, and that makes them bad residues to sit inside helices or strands,
but good residues for β-turns. Etcetera.