Monday, March 19, 2007
Since its original proposition by HJ Jeffrey in the early 90's, Chaos Game Representation (CGR) has been intriguing a small number of researchers as a generalized representation of Markovian processes. After an initial period where the graphical appeal of the technique took precedence over its deeper properties, it was shown that indeed Markov transition is a special case of CGR rather than the opposite as proposed soon after the original publication (see that link for an historic overview of the first decade of CGR work). That conclusion creates the opportunity to revisit basic concepts of pattern recognition in symbolic sequences that are particularly relevant to the analysis of Biological sequences. That domain is particularly interesting because it is well understood that Biological function has a dependency on sequence that takes place, simultaneously, at multiple scales. For example, single nucleotide mutations can have a drastic effect in gene expression and so does Chromatin organization. GENECHAOS.ORG (this resource) is created as a collaboratorium open to researchers in this field.