The benefits of a pre-computed amino acid structure database in quantum chemical geometry optimizations of β-turns of peptides

Energetically most favored conformations of individual amino acids can be used as building blocks, using a recently proposed system for logical numbering of atomic nuclei, to enhance the ab initio quantum chemical geometry optimizations of oligopepides. The structures of amino acid residues previously optimized at the RHF/3-21G and the RHF/6-31G(d) levels of theory were used for input generation and the structures of Ac-Asn-Pro-Gly-Gln-NH₂ and Ac-Val-Pro-D-Ala-His-NH₂, in their experimentally proposed type I and type II β-turn conformations, respectively, were optimized by using the B3LYP and the B3PW91 density functionals on the 6-31G(d) basis set. Input generation, by extracting the internal coordinates of residues from their previously optimized structures, resulted in quicker convergence of optimization showing no significant influence of the basis set on which the pre-optimization was carried out. However, efficiency of the two density functional theory methods differed.