Institutions | About Us | Help | Gaeilge
rian logo


Mark
Go Back
HIV modelling - parallel implementation strategies
Perrin, Dimitri; Ruskin, Heather J.; Crane, Martin
We report on the development of a model to understand why the range of experience with respect to HIV infection is so diverse, especially with respect to the latency period. To investigate this, an agent-based approach is used to extract highlevel behaviour which cannot be described analytically from the set of interaction rules at the cellular level. A network of independent matrices mimics the chain of lymph nodes. Dealing with massively multi-agent systems requires major computational effort. However, parallelisation methods are a natural consequence and advantage of the multi-agent approach and, using the MPI library, are here implemented, tested and optimized. Our current focus is on the various implementations of the data transfer across the network. Three communications strategies are proposed and tested, showing that the most efficient approach is communication based on the natural lymph-network connectivity.
Keyword(s): Computational complexity; Mathematical models; Diseases; Immunology; Virology; HIV; immune modelling; MPI; parallelisation
Publication Date:
2006
Type: Other
Peer-Reviewed: Unknown
Language(s): English
Institution: Dublin City University
Citation(s): Perrin, Dimitri, Ruskin, Heather J. ORCID: 0000-0001-7101-2242 <https://orcid.org/0000-0001-7101-2242> and Crane, Martin ORCID: 0000-0001-7598-3126 <https://orcid.org/0000-0001-7598-3126> (2006) HIV modelling - parallel implementation strategies. In: Third International Conference on Cluster and Grip Computing Systems, 24-26 November 2006, Venice, Italy.
File Format(s): application/pdf
Related Link(s): http://doras.dcu.ie/9/1/CGCS_06_-_Dimitri_Perrin_-_HIV_Modelling_-_Parallel_Implementation_Strategies%28Paper_code_20045%29.pdf,
http://www.enformatika.org/cgcs06/
First Indexed: 2009-11-05 02:00:10 Last Updated: 2019-02-09 06:05:09