Using long fragments to reconstruct RNA backbones

This study addresses to what extent RNA backbones consist of long, recurring fragments. A dynamic programming algorithm is presented that finds the smallest set of fragments that are required to reconstruct the backbone of a known macromolecule within a given root mean square bound. This algorithm is guaranteed to find a global minimum. The program has been used for reconstructing the backbones of several known RNA structures, using fragments taken from ribosomal RNA. Each of these target structures could be approximated within a root mean square deviation of 1.0 Aring using fragments with an average length of more than eight nucleotides.