Figure 1. Scheme to obtain sequences meeting the criteria established to be considered as crustins.

The pipeline was as follows: 1. Crustin or carcinin sequences were searched in the GenBank; 2. repeated or incompleted sequences and those containing <12 Cys or <2 CC pairs were excluded; 3. additional sequences similar to any listed crustin were searched by BlastP, and those having an E value < =9.9E−20 were selected; 4. sequences without signal peptide or initial Met or containing <12 Cys or <2 CC pairs are considered incomplete and excluded; 5. only 233 sequences having the crustin signature (Cys-rich region and 4-DSC domain) were selected; 6. crustin signatures were used as the first classification criterium; 7. grouped by their Cys-rich motif, the most conserved part of the signature; 8. cladograms with the most abundant motifs were done; 9. crustin types were assigned based on the Gly-rich region; and 10. a table of distances was used for comparison, and group crustins with similar signatures.