PreRkTAG: Prediction of RNA Knotted Structures Using Tree Adjoining Grammars

-

Background: RNA molecules play many important regulatory, catalytic and structuralroles in the cell, and RNA secondary structure prediction with pseudoknots is one themost important problems in biology. An RNA pseudoknot is an element of the RNA secondary structure in which bases of a single-stranded loop pair with complementary basesoutside the loop. Modeling these nested structures (pseudoknots) causes numerous computational diffilties and so it has been generally neglected in RNA structure predictionalgorithms.Objectives: In this study, we present a new heuristic algorithm for the Prediction of RNAKnotted structures using Tree Adjoining Grammars (named PreRKTAG).Materials and Methods: For a given RNA sequence, PreRKTAG uses a genetic algorithm ontree adjoining grammars to propose a structure with minimum thermodynamic energy.The genetic algorithm employs a subclass of tree adjoining grammars as individuals bywhich the secondary structure of RNAs are modeled. Upon the tree adjoining grammars,new crossover and mutation operations were designed.The finess function is defied according to the RNA thermodynamic energy function, which causes the algorithm convergence to be a stable structure.Results: The applicability of our algorithm is demonstrated by comparing its iresults withthree well-known RNA secondary structure prediction algorithms that support crossedstructures.Conclusions: We performed our comparison on a set of RNA sequences from the RNAsePdatabase, where the outcomes show effiency and practicality of the proposed algorithm.