Dynamic Programming Algorithms and Grammatical Modeling for Protein Beta-Sheet Prediction

Abstract

Abstract

Protein structure prediction is one major task in computational biology, and various methods have been proposed for secondary structure prediction. In particular, it is a challenge to predict β-sheet structures since they range over several discontinuous regions in an amino acid sequence. In this article, we propose dynamic programming algorithms for several kinds of antiparallel β-sheets, where the proposed approach can be extended for more general classes of β-sheets. Experimental results for real data show that our prediction algorithms have good performance in accuracy. We also show a relation between the proposed algorithms and an existing grammar-based method, and present another grammatical modeling. Furthermore, we prove that prediction of planar β-sheet structures is NP-hard.