ON THE LEVEL DISTRIBUTION OF PERIODIC AND NONPERIODIC COLLAGEN MODELS

Citation
G. Nemethy et al., ON THE LEVEL DISTRIBUTION OF PERIODIC AND NONPERIODIC COLLAGEN MODELS, Journal of molecular structure. Theochem, 337(2), 1995, pp. 103-107
Citations number
25
Categorie Soggetti
Chemistry Physical
ISSN journal
01661280
Volume
337
Issue
2
Year of publication
1995
Pages
103 - 107
Database
ISI
SICI code
0166-1280(1995)337:2<103:OTLDOP>2.0.ZU;2-L
Abstract
The negative factor counting method in its matrix block form is used t o calculate the total electronic density of states (DOS) of different collagen models. The computations have been executed for periodic and non-periodic eight-component collagen models consisting of (glycine, p roline, hydroxypoline and alanine) together with glutamine, arginine, leucine and aspartic acid. The Fock and overlap matrices of the whole systems (containing 1014 amino acid residues) have been built up from ab initio Hartree-Fock dimer calculations (29 different dimers). In co nstructing these matrices both for the periodic and non-periodic cases , the well known sequence regularities of collagen have been taken int o account. The conformation of the chain has been taken from a collage n-like model triple helix, Nemethy and Scheraga having determined the conformation of this triple helix with the aid of empirical potential energy calculations. The DOS histogram of periodic (Gly-X-Y) triplets (similar to the previous six-component case) shows only very narrow pe aks in both the valence band and conduction band regions. As a next st ep the 20 possible different triplets have been arranged in a (composi tion restricted) random manner (338 triplets). The DOS histogram shows in these cases essentially broader allowed energy regions and a much lower number of very narrow peaks than in the previous six-component c ase. Therefore we can conclude that if one were able to calculate the DOS of real collagen (with 20 different amino acid residues), the DOS histograms, the Anderson localization and the hopping frequencies woul d be very similar to those in random polypeptides.