High-throughput method for the prediction of low-resolution, three-dimensional RNA structures.

Abstract

Our aim is the high-throughput prediction of three-dimensional RNA structures at low resolution to answer questions posed by the RNA molecular biology. The strategy presented here is based on a computer program, called 3D-RNApredict, which implements and converts various structural data (RNA secondary structure, RNA motifs coordinates, experimental data) to create the input to the CYANA program (torsion-angle dynamics algorithm, TAD) which provides a fast engine for the 3D RNA structure calculation. The resultant RNA structure is refined using the X-PLOR program. The approach is amenable to automatization. Current predictions involve HIV-2 leader RNA domains and the sets of bacterial ncRNA.