Pattern-matching search of DNA sequences using logic grammars

An approach to DNA sequence analysis founded on principles and techniques of computational linguistics is described. By specifying elements of sequences with grammars, it is possible to perform search and analysis by parsing. Such grammars offer greater representational power than current methods in this domain by allowing an abstract, hierarchical, and declarative system for describing features of interest. The authors have imported artificial intelligence (AI) methods to create a highly flexible parsing system, using the logic-based definite clause grammar formalism. they show how AI techniques such as chart parsing, backtracking control and imperfect matching can be used to great advantage in this domain. Examples are given of parses for a wide variety of viral genomes and globin genes