| Abstract: | Compact polymers such as proteins obtain their unique conformation by appropriate nonbonded interactions among their monomer residues. Innumerable nonnative compact conformations are also possible, and it is essential to distinguish the native from the nonnative conformations. Toward this goal we have used graph‐theoretic methods to classify polymer structures formed by noncovalent interactions. All compact structures on a 4×4 two‐dimensional lattice and a few conformations on 3×3×3 cubic lattice have been investigated. The 69 compact conformations in 4×4 two‐dimensional lattice are classified into 12 groups based on the highest eigenvalue and eigenvector. The complex graphs obtained for polymers in a 3×3×3 lattice space are analyzed. Their eigenvalues and eigenvector components are correlated with the branching structure and the center of the graph. The method has application in classifying real polymers such as proteins into their substructures, cluster, and domains. ©1999 John Wiley & Sons, Inc. Int J Quant Chem 71: 349–356, 1999 |
