Predicting E. Coli promoters using formal grammars

Ever since the structure of the DNA was discovered, linguistics has been part of molecular biology [13]. Grammatical linguistics is a powerful method to express information and describe its structure. It can be used to express transcribed in DNAs. Most formal grammar applications on DNAs are based on Searls DNA parsing approach using Prolog-based Definite Clause Grammars (DFG) [11]. Extensions of this approach include String Variable Grammar [6] and Basic Gene Grammars [5]. This paper presents a novel approach by parsing Escherichia Coli (E. Coli) promoter sequences using a Context-Free Grammar (CFG). The approach takes advantage of an error correcting parsing algorithm introduced by Rajasekaran and Nicolae [1]. The idea is to derive a grammar for known promoter regions and then modify this grammar to tolerate errors. The resulting cover grammar can then be employed to recognize promoter regions. Introducing probabilities in the production rules can further extend the cover grammar. Please note that in this paper we introduce this novel paradigm. In our future work we will implement this approach and test it on various datasets.