Figure 4

Comparisons of amino acid compositions, secondary structure types and relative solvent accessibilities of indel residues in "all indels", "naturally occurring indels" and reference datasets. Relative frequencies of 20 amino acids, frequencies of secondary structure types (helix, strand, coil, and disordered), and relative solvent accessibilities (buried: ≤7%, intermediate: >7% and ≤37%, exposed: >37%) are shown in A, B and C respectively. The one-letter code for amino acids is used. "Background" data for amino acid frequencies, secondary structure types and solvent accessibilities are calculated from a dataset of 4731 non-redundant protein structures (See Methods). "Natural" represents an indel dataset without engineered indels. "All" indel dataset includes both engineered and natural indel sequences.