RNAMotifContrast: a method to discover and visualize RNA structural motif subfamilies

Understanding the 3D structural properties of RNAs will play a critical role in identifying their functional characteristics and designing new RNAs for RNA-based therapeutics and nanotechnology. While several existing computational methods can help in the analysis of RNA properties by recognizing structural motifs, they do not provide the means to compare and contrast those motifs extensively. We have developed a new method, RNAMotifContrast, which focuses on analyzing the similarities and variations of RNA structural motif characteristics. In this method, a graph is formed to represent the similarities among motifs, and a new traversal algorithm is applied to generate visualizations of their structural properties. Analyzing the structural features among motifs, we have recognized and generalized the concept of motif subfamilies. To asses its effectiveness, we have applied RNAMotifContrast on a dataset of known RNA structural motif families. From the results, we observed that the derived subfamilies possess unique structural variations while holding standard features of the families. Overall, the visualization approach of this method presents a new perspective to observe the relation among motifs more closely, and the discovered subfamilies provide opportunities to achieve valuable insights into RNA's diverse roles.