A New Method of RNA Secondary Structure Prediction Based on Convolutional Neural Network and Dynamic Programming

Abstract

In recent years, obtaining RNA secondary structure information has played an important role in RNA and gene function research. Although some RNA secondary structures can be gained experimentally, in most cases, efficient, and accurate computational methods are still needed to predict RNA secondary structure. Current RNA secondary structure prediction methods are mainly based on the minimum free energy algorithm, which finds the optimal folding state of RNA in vivo using an iterative method to meet the minimum energy or other constraints. However, due to the complexity of biotic environment, a true RNA structure always keeps the balance of biological potential energy status, rather than the optimal folding status that meets the minimum energy. For short sequence RNA its equilibrium energy status for the RNA folding organism is close to the minimum free energy status; therefore, the minimum free energy algorithm for predicting RNA secondary structure has higher accuracy. Nevertheless, in a longer sequence RNA, constant folding causes its biopotential energy balance to deviate far from the minimum free energy status. This deviation is because of its complex structure and results in a serious decline in the prediction accuracy of its secondary structure. In this paper, we propose a novel RNA secondary structure prediction algorithm using a convolutional neural network model combined with a dynamic programming method to improve the accuracy with large-scale RNA sequence and structure data. We analyze current experimental RNA sequences and structure data to construct a deep convolutional network model, and then we extract implicit features of an effective classification from large-scale data to predict the pairing probability of each base in an RNA sequence. For the obtained probabilities of RNA sequence base pairing, an enhanced dynamic programming method is applied to obtain the optimal RNA secondary structure. Results indicate that our proposed method is superior to the common RNA secondary structure prediction algorithms in predicting three benchmark RNA families. Based on the characteristics of deep learning algorithm, it can be inferred that the method proposed in this paper has a 30% higher prediction success rate when compared with other algorithms, which will be needed as the amount of real RNA structure data increases in the future.

Characterization of lncRNA-Perturbed TLR-Signaling Network Identifies Novel lncRNA Prognostic Biomarkers in Colorectal Cancer

“…The resulting lncRNA-perturbed TLR-signaling network contained 280 lncRNAs and 122 TLR genes, implying that abnormal expression of these 280 lncRNAs might perturb the TLRsignaling network which contributed CRC development and progression. A large number of studies have indicated that lncRNAs are an essential component in cancer biology by acting as oncogenic and tumor-suppressive factors, and have become emerging and suitable diagnostic and prognostic biomarkers or therapeutic targets in various cancers including CRC (Gibb et al, 2011;Cheetham et al, 2013;Kornienko et al, 2013;Zhu et al, 2015;Zhou et al, 2015bZhou et al, , 2017Zhou et al, , 2018bZhou et al, , 2020Zhong et al, 2018;Wang et al, 2019;Zhang et al, 2019;Sun et al, 2020). Therefore, we further explored the prognostic value of these 280 TLRrelated lncRNAs in CRC patients.…”

Collaborative Learning of Estimation of Distribution Algorithm for RNA secondary structure prediction

“…Until now, there has not been a biological RNA method that can correctly predict a true RNA secondary structure in large quantities. Thus, effective computational prediction algorithms are still needed [6].…”

“…It is excellent, but not actually optimal in the real world [9]. That is why current RNA secondary structure prediction methods are mainly based on the minimum free energy algorithm [6] and remain essential tools for RNA structural biology [10].…”

“…DpacoRNA [9] applied a parallel ant colony optimization strategy combined with a bi-directional LSTM recurrent neural network. CDPfold [6] uses a convolutional neural network model combined with a dynamic programming method to improve the accuracy for large-scale RNA sequences. Moreover, some studies use many objective functions.…”

Refining RNA solution structures with the integrative use of label-free paramagnetic relaxation enhancement NMR

“…But unlike proteins, the RNA folding is basically hierarchical (Herschlag et al 2018). As a result, though there are programs to predict RNA secondary structures regarding the patterns of base-pairing (Bellaousov et al 2013;Parisien and Major 2008;Zhang et al 2019;Zuker 2003), it remains difficult to predict RNA secondary structures with complex topologies (Schlick and Pyle 2017;Zhao et al 2018).…”