A novel text modeling approach for structural comparison and alignment of biomolecules

Abstract

Within this paper, a novel strategy for structural alignment of proteins based on text modeling techniques is introduced. The method summarizes the protein secondary and tertiary structure in two textual sequences. The first sequence is used to initial superposition of secondary structure elements and the second sequence is employed to align the 3D-structure of two compared structures. The comparison technique used by the method has been inspired from computational linguistics for analysing and quantifying textual sequences. In this strategy, the cross-entropy measure over n-gram models is used to capture regularities between sequences of protein structures. The performance of the method is evaluated and compared with CE and SSM methods. The results of the experiments reported here provide evidence for the preference and applicability of the new approach in terms of efficiency and effectiveness.