An RNA Scoring Function For Tertiary Structure Prediction Based On Multi-layer Neural Networks

Ribonucleic acids(RNA)are essential constituents of living organism.They play important roles in coding,decoding,regulation,and expression of genes.The knowledge of their three-dimensional structures are generally needed for people understanding their functions.At the present stage,X-ray crystallography and nuclear magnetic resonance spectroscopy have been employed to determine RNAs' 3D structures.Such experiments,however,are costly,time-consuming and also technically challenging.These facts urge the computational methods to be developed.A good scoring function is necessary for ab inito prediction of RNA tertiary structures.In this study,we explored the power of a machine learning based approach as a scoring function.Compared with the traditional scoring functions,the present approach is more flexible in incorporating different kinds of features;it is also free of the difficult problem of choosing the reference state.Two multi-layer neural networks were constructed and trained.They took RNA a structural candidate as input and then output its likeness score that evaluates the likeness of the candidate to the native structure.The first network was working at the coarse-grained level of RNA structures,while the second at the all-atom level.We also built an RNA database and split it into the training,validation,and testing sets,containing 322,70,and 70 RNAs,respectively.Each RNA was accompanied with 300 decoys generated by high-temperature molecular dynamics simulations.The networks were trained on the training set and then optimized with an early-stop strategy,based on the loss of the validation set.We then tested the performance of the networks on the testing set.The results were found to be consistently better than a recent knowledge-based all-atom potential.This dissertation is organized as follows:Chapter one is a brief review of relevant background information related to RNA and neural network.In Chapter two,we describe our RNA scoring function for tertiary structure prediction based on multi-layer neural networks and the main results.Chapter three is a summary of this dissertation.