Integrating end-to-end learning with deep geometrical potentials for ab initio RNA structure prediction

Abstract

RNAs are fundamental in living cells and perform critical functions determined by the tertiary architectures. However, accurate modeling of 3D RNA structure remains a challenging problem. Here we present a novel method, DRfold, to predict RNA tertiary structures by simultaneous learning of local frame rotations and geometric restraints from experimentally solved RNA structures, where the learned knowledge is converted into a hybrid energy potential to guide subsequent RNA structure constructions. The method significantly outperforms previous approaches by >75.6% in TM-score on a nonredundant dataset containing recently released structures. Detailed analyses showed that the major contribution to the improvements arise from the deep end-to-end learning supervised with the atom coordinates and the composite energy function integrating complementary information from geometry restraints and end-to-end learning models. The open-source DRfold program allows large-scale application of high-resolution RNA structure modeling and can be further improved with future release of RNA structure databases.