Comparison of Two Folding Functions for RNA Secondary Structure Design

Using computational programs to predict Ribonucleic acid (RNA) sequence with desired properties has been extensively developed in the past few years. The RNA inverse folding is an NP-Hard problem that describes the process of finding an RNA primary structure that will fold into a given RNA secondary structure. With new advances in the new fields of synthetic biology and RNA nanostructures, there is a growing interest in studying RNA inverse folding. In most of the RNA inverse folding algorithms, in order to evaluate the fitness of a given primary structure, a computationally expensive folding operation is required. In this paper, we incorporated two different folding algorithms in our Simulated Annealing RNA Design (SIMARD) algorithm while employing the Dynamic Exploration Strategy (DES). We employed ViennaRNA and pKiss as our folding operations. Our findings show a significant potential to use both folding operations for our RNA design algorithm. We found that using ViennaRNA tended to produce a more accurate structure, and pKiss tended to produce structures with lower thermodynamic free energy.