This report describes in detail version 0.1 of a specification for a nonstandard computational model that emulates the reactivity of enzymes in bacterial cells. The model is an artificial chemistry in which the unique properties of each molecular species is encoded as a sequence of symbols. The model has a very simple cell-level representation consisting of a mixing volume and mixing equation. There is no distinction between genotype and phenotype, or data and program. Randomly selected molecules are subject to a stochastic binding test, in which their identifying sequences are subjected to a complementary alignment process. If the bind is successful, the molecules are combined to form a computational entity to run the reaction between them. At this point the sequence acts as a microprogram. These microprograms can perform a range of tasks, much as a chemical reaction can. Our first chemistry within this model solves the problem of designing a chemical "replicase", capable of creating copies of itself such that it can replenish a population of molecules that are subject to a decay process. This report gives detail to the model, describes the replicase molecule and its function, and shows an "invasion when rare" experiment.
Full report : PDF 727K | a full specification of the version of stringmol used in the ECAL09 paper
@techreport(SS-YCS-457, author = "Simon Hickinbotham and Edward Clark and Susan Stepney and Tim Clarke and Adam Nellis and Mungo Pay and Peter Young", title = "Specification of the stringmol chemical programming language version 0.1", institution = "Department of Computer Science, University of York", number = "YCS-2010-457", year = 2010 )