DotKnot: pseudoknot prediction using the probability dot plot under a refined energy model

Sperschneider, Jana; Datta, Amitava

DotKnot predicts the class of recursive H-type pseudoknots where one of the pseudoknot stems can be interrupted by bulge or internal loops. All of the three pseudoknot loops are allowed to form internal secondary structures. Using the set of structural elements derived from the probability dot plot as a construction kit, we could predict a broad class of pseudoknots, including kissing hairpins, in an efficient manner. However, our knowledge about energy parameters and folding mechanisms become the limiting factor for a pseudoknot search tool such as DotKnot. Given an RNA sequence, DotKnot returns only the (possibly empty) set of detected pseudoknots. These detected pseudoknots can subsequently be verified using laboratory techniques or comparative information. The remaining non-crossing sequence can then be folded in Oðn 3 Þ time and Oðn 2 Þ space using state-of-the-art RNA secondary structure prediction algorithms to obtain a global folding.