Abstract
36 mutants of the Sulfolobus solfataricus amidase were analyzed by comparing biochemical data to structural data obtained by a learning machine. The analysis shows that beside well known catalytic residues, amino acid residues Arg197, Lys209 and Asp228 are important for the catalytic activity of the signature thermophilic amidase.
Keywords: Functional residues, amidase, machine learning, SVM, FRAP