Kenway, O.A.; (2010) Molecular dynamics simulations of complex systems including HIV-1 protease. Doctoral thesis , UCL (University College London).

Preview

PDF 19485.pdf Download (10MB)

Abstract

Advances in supercomputer architectures have resulted in a situation where many scienti�fic codes are used on systems whose performance characteristics di�ffer considerably from the platform they were developed and optimised for. This is particularly apparent in the realm of Grid computing, where new technologies such as MPIg allow researchers to connect geographically disparate resources together into virtual parallel machines. Finding ways to exploit these new resources efficiently is necessary both to extract the maximum bene�fit from them, and to provide the enticing possibility of enabling new science. In this thesis, an existing general purpose molecular dynamics code (LAMMPS) is extended to allow it to perform more efficiently in a geographically distributed Grid environment showing considerable performance gains as a result. The technique of replica exchange molecular dynamics is discussed along with its applicability to the Grid model and its bene�fits with respect to increasing sampling of configurational space. The dynamics of two sub-structures of the HIV-1 protease (known as the flaps) are investigated using replica exchange molecular dynamics in LAMMPS showing considerable movement that would have been difficult to investigate by traditional methods. To complement this, a study was carried out investigating the use of computational tools to calculate binding affinity between HIV-1 protease mutants and the drug lopinavir in comparison with results derived experimentally by other research groups. The results demonstrate some promise for computational methods in helping to determine the most eff�ective course of treatment for patients in the future.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item