Computational analysis of intramolecular interactions in proteins.
Ian K. McDonald · OpenGrey (Institut de l'Information Scientifique et Technique) · 1996
Hydrogen bonds are one of the dominant forms of atomic interaction within proteins and are studied extensively here. The thesis established a non-homologous dataset extracted from the Brookhaven Protein Databank and hydrogen-bond criteria based on geometrical distributions. It analyses the frequency and geometry with which potential hydrogen bond donors and acceptors are satisfied in protein molecules. An Atlas of Side-Chain and Main-Chain Hydrogen Bonding is created and trends between different donor and acceptor types are drawn out and examined closely. For high resolution structures, 9.5% and 5.1% of buried main-chain nitrogen and oxygen atoms, respectively, fail to hydrogen bond under standard criteria. The main-chain atoms dominate protein hydrogen bonding and are examined in particular detail. This includes examining the environment of unsatisfied atoms and correlations between the percentages of unsatisfied atoms and measures of structural quality. Based on the observation that almost all buried donors and acceptors are satisfied, a simple algorithm is developed to compare the alternative conformations of Asn, Gln and His side-chains, the orientations of which are ambiguous in purely X-ray crystal structures, and where possible identify the most favourable.