Protein folding with stochastic L-systems

Protein molecules adopt a specific global 3D structure in order to carry out their biological function. To achieve this native state a newly formed protein molecule has to fold. The folding process and the final fold are both determined by the sequence of amino acids making up the protein chain. It is not currently possible to predict the conformation of the native state from the amino acid sequence alone and the protein folding process is still not fully understood. We are using L-systems, sets of rewriting rules, to model the folding of protein-like structures. Models of protein folding vary in complexity and the amount of prior knowledge they contain on existing native protein structures. In a previous paper we presented a method of using open L-systems to model the folding of protein-like structures using physics-based rewriting rules. Here we present an L-systems model of protein folding that uses knowledge-based rewriting rules and stochastic L-systems.