Calculation of protein tertiary structure

We describe a method for calculating the tertiary structure of proteins given their amino acid sequence. The algorithm involves locally minimizing an energylike expression as a function of the Cartesian co-ordinates of the C_β of all residues. Although the approximation to the true polypeptide geometry and conformational energies is extremely approximate, quite respectable results have been obtained for the small proteins rubredoxin and trypsin inhibitor, where the root mean square errors are as low as 4.0 Å and 4.7 Å, respectively.