Visual Data Mining of RNA Secondary Structure and Folding Pathways as Determined by the Massively Parallel Genetic Algorithm

RNA folding pathways are proving to be quite important in the determination of RNA function. Studies indicate that RNA may enter intermediate and multiple conformational states that are key to its functionality. These states may have a significant impact on gene expression and molecular function. It is known that the biologically functional states of RNA molecules may not correspond to their minimum energy state, that kinetic barriers may exist that trap the molecule in a local minimum, that folding often occurs during transcription, and that cases exist in which a molecule will transition between one or more functional conformations. Thus, methods for simulating the folding pathway and dynamic behavior of an RNA molecule are important for the prediction of RNA structure and its associated functions. We have developed several visual data mining techniques associated with a massively parallel genetic algorithm for RNA structure prediction, as well as with STRUCTURELAB, our RNA/DNA structure analysis workbench. These methodologies are used to determine the significant intermediate and final structures associated with RNA folding. Since the genetic algorithm is essentially stochastic, multiple runs are required. The visualization procedures used give significant feedback concerning the characteristics of the folding runs. This feedback encompasses: interpretation of results from individual genetic algorithm runs that are based on population consensus or best fit structures, this includes the discovery of transition states in the folding process; final results of individual runs; and the interpretation of genetic algorithm results from multiple RNA sequences from the same family to identify common structural elements across the different sequences. In addition, fitness maps as well as results derived from different population sizes are used. The combination of the visualization techniques as well as other methodologies embedded within the STRUCTURELAB and genetic algorithm environments help to determine the overall picture representing the folding pathway or final structure(s) of a given RNA sequence. This paper will describe several of these techniques and show how they are used to help solve this very highly combinatoric problem.