An unexpected protein-fold evolution via natural selection was found in the crystal structure of congerin I, a congar eel galectin. This particular example suggests that the relationship among protein-folds should be studied not only by stereochemistry but also by phylogenetics. How can we predict entire fold models exist in the protein-fold space and their relationships? The outline and the preliminary results from an attempt for mapping the protein-fold space by a computer simulation are described.
