Semi-supervised segmentation of RNA 3D structures using density-based clustering

  Non-coding RNAs, especially the long ones, are likely to contain structurally compact regions, reminiscent of the 3D domains found in proteins. Knowing the domain composition of RNA is necessary for a divide-and-conquer strategy to solve the RNA folding problem. The RNA3DClust tool can segment RNA molecules into such substructures, following a clustering-based approach.