Damer, Bruce, Peter Newman, Richard Gordon, Tom Barbalet, David Deamer, and Ryan Norkus. “The EvoGrid-A Framework for Distributed Artificial Chemistry Cameo Simulations Supporting Computational Origins of Life Endeavors.” In ALIFE , pp. 73-79. 2010.
URL1 URL2
The Evolution Grid, or EvoGrid is a computer simulation framework for distributed artificial chemistry (AC) supporting computational origins of life (COoL) research. The EvoGrid consists of a number of small experiments running on short time scales pruned by aggressive tree-branching searches supported by random parametric re-seeding and temporal back-tracking. The EvoGrid is designed to converge upon the observation of “cameo” simulations of key pre-biotic or simple biological structures or behaviors. These cameo simulations can then inform and feed larger AC simulations operating over biologically relevant time scales. In addition, the framework is designed to plug into a heterogeneous set of engines ranging from high fidelity molecular dynamics (MD) to more abstract AC techniques on the same set of data. The EvoGrid also provides shared web-based simulation management services and uniform, open standards for execution, storage and data analysis. We conclude by describing the first prototype implementation of the EvoGrid, early results, next steps and open questions in this and other COoL endeavors.