RNA Structure Prediction Research Articles

Machine learning modeling of RNA structures: methods, challenges and future perspectives.

The three-dimensional structure of RNA molecules plays a critical role in a wide range of cellular processes encompassing functions from riboswitches to epigenetic regulation. These RNA structures are incredibly dynamic and can indeed be described aptly as an ensemble of structures that shifts in distribution depending on different cellular conditions. Thus, the computational prediction of RNA structure poses a unique challenge, even as computational protein folding has seen great advances. In this review, we focus on a variety of machine learning-based methods that have been developed to predict RNA molecules' secondary structure, as well as more complex tertiary structures. We survey commonly used modeling strategies, and how many are inspired by or incorporate thermodynamic principles. We discuss the shortcomings that various design decisions entail and propose future directions that could build off these methods to yield more robust, accurate RNA structure predictions.

Using 3dRNA/DNA for RNA and DNA 3D Structure Prediction and Evaluation.

This article describes how to use 3dRNA/DNA to predict RNA and DNA 3D structures from their sequences and secondary (2D) structures. The predicted RNA and DNA 3D structures can be used to predict or understand their functions and can also be used to find the interactions between the RNA/DNA and other molecules. © 2023 Wiley Periodicals LLC. Basic Protocol: RNA/DNA 3D structure prediction using 3dRNA/DNA web server Alternate Protocol: Using 3dRNA/DNA and 3dRNAscore locally Support Protocol: Installing NSP.

Syntactic Pattern Recognition for the Prediction of L-Type Pseudoknots in RNA

The observation and analysis of RNA molecules have proved crucial for the understanding of various processes in nature. Scientists have mined knowledge and drawn conclusions using experimental methods for decades. Leveraging advanced computational methods in recent years has led to fast and more accurate results in all areas of interest. One highly challenging task, in terms of RNA analysis, is the prediction of its structure, which provides valuable information about how it transforms and operates numerous significant tasks in organisms. In this paper, we focus on the prediction of the 2-D or secondary structure of RNA, specifically, on a rare but yet complex type of pseudoknot, the L-type pseudoknot, extending our previous framework specialized for H-type pseudoknots. We propose a grammar-based framework that predicts all possible L-type pseudoknots of a sequence in a reasonable response time, leveraging also the advantages of core biological principles, such as maximum base pairs and minimum free energy. In order to evaluate the effectiveness of our methodology, we assessed four performance metrics: precision; recall; Matthews correlation coefficient (MCC); and F1-score, which is the harmonic mean of precision and recall. Our methodology outperformed the other three well known methods in terms of Precision, with a score of 0.844, while other methodologies scored 0.500, 0.333, and 0.308. Regarding the F1-score, our platform scored 0.671, while other methodologies scored 0.661, 0.449, and 0.449. The proposed methodology surpassed all methods in terms of the MCC metric, achieving a score of 0.521. The proposed method was added to our RNA toolset, which aims to enhance the capabilities of biologists in the prediction of RNA motifs, including pseudoknots, and holds the potential to be applied in a multitude of biological domains, including gene therapy, drug design, and comprehending RNA functionality. Furthermore, the suggested approach can be employed in conjunction with other methodologies to enhance the precision of RNA structure prediction.

Identification of metal ion-binding sites in RNA structures using deep learning method.

Metal ion is an indispensable factor for the proper folding, structural stability and functioning of RNA molecules. However, it is very difficult for experimental methods to detect them in RNAs. With the increase of experimentally resolved RNA structures, it becomes possible to identify the metal ion-binding sites in RNA structures through in-silico methods. Here, we propose an approach called Metal3DRNA to identify the binding sites of the most common metal ions (Mg2+, Na+ and K+) in RNA structures by using a three-dimensional convolutional neural network model. The negative samples, screened out based on the analysis for binding surroundings of metal ions, are more like positive ones than the randomly selected ones, which are beneficial to a powerful predictor construction. The microenvironments of the spatial distributions of C, O, N and P atoms around a sample are extracted as features. Metal3DRNA shows a promising prediction power, generally surpassing the state-of-the-art methods FEATURE and MetalionRNA. Finally, utilizing the visualization method, we inspect the contributions of nucleotide atoms to the classification in several cases, which provides a visualization that helps to comprehend the model. The method will be helpful for RNA structure prediction and dynamics simulation study. Availability and implementation: The source code is available at https://github.com/ChunhuaLiLab/Metal3DRNA.

Pseudoknots in RNA Structure Prediction.

RNA molecules play active roles in the cell and are important for numerous applications in biotechnology and medicine. The function of an RNA molecule stems from its structure. RNA structure determination is time consuming, challenging, and expensive using experimental methods. Thus, much research has been directed at RNA structure prediction through computational means. Many of these methods focus primarily on the secondary structure of the molecule, ignoring the possibility of pseudoknotted structures. However, pseudoknots are known to play functional roles in many RNA molecules or in their method of interaction with other molecules. Improving the accuracy and efficiency of computational methods that predict pseudoknots is an ongoing challenge for single RNA molecules, RNA-RNA interactions, and RNA-protein interactions. To improve the accuracy of prediction, many methods focus on specific applications while restricting the length and the class of the pseudoknotted structures they can identify. In recent years, computational methods for structure prediction have begun to catch up with the impressive developments seen in biotechnology. Here, we provide a non-comprehensive overview of available pseudoknot prediction methods and their best-use cases. © 2023 Wiley Periodicals LLC.

Predicting 3D RNA structure from solely the nucleotide sequence using Euclidean distance neural networks.

Fast and accurate 3D RNA structure prediction remains a major challenge in structural biology, primarily due to the size and flexibility of RNA molecules, as well as lack of diverse experimentally determined structures of RNA molecules. Unlike DNA structure, RNA structure is far less constrained by base pair hydrogen bonding, resulting in an explosion of potential stable states. Here, we propose a convolutional neural network which predicts all pairwise distances between residues in an RNA, using a recently described smooth parametrization of Euclidean distance matrices. We achieve high accuracy predictions on RNAs up to 100 nucleotides in length in fractions of a second, a factor of 107 faster than existing molecular dynamics-based methods. We also convert our coarse-grained machine learning output into an all-atom model using discrete molecular dynamics with constraints. Our proposed computational pipeline accurately predicts all-atom RNA models solely from the nucleotide sequence.

A Conserved Stem-Loop Structure within ORF5 Is a Frequent Recombination Hotspot for Porcine Reproductive and Respiratory Syndrome Virus 1 (PRRSV-1) with a Particular Modified Live Virus (MLV) Strain

The emergence of recombinant PRRSV strains has been observed for more than a decade. These recombinant viruses are characterized by a genome that contains genetic material from at least two different parental strains. Due to the advanced sequencing techniques and a growing number of data bank entries, the role of PRRSV recombinants has become increasingly important since they are sometimes associated with clinical outbreaks. Chimeric viruses observed more recently are products of PRRSV wild-type and vaccine strains. Here, we report on three PRRSV-1 isolates from geographically distant farms with differing clinical manifestations. A sequencing and recombination analysis revealed that these strains are crossovers between different wild-type strains and the same modified live virus vaccine strain. Interestingly, the recombination breakpoint of all analyzed isolates appears at the beginning of open reading frame 5 (ORF5). RNA structure predictions indicate a conserved stem loop in close proximity to the recombination hotspot, which is a plausible cause of a polymerase template switch during RNA replication. Further research into the mechanisms of the stem loop is needed to help understand the PRRSV recombination process and the role of MLVs as parental strains.

Genome-Wide RNA Secondary Structure Prediction.

The information of RNA secondary structure has been widely applied to the inference of RNA function. However, a classical prediction method is not feasible to long RNAs such as mRNA due to the problems of computational time and numerical errors. To overcome those problems, sliding window methods have been applied while their results are not directly comparable to global RNA structure prediction. In this chapter, we introduce ParasoR, a method designed for parallel computation of genome-wide RNA secondary structures. To enable genome-wide prediction,ParasoR distributes dynamic programming (DP) matrices required for structure prediction to multiple computational nodes. Using the database of not the original DP variable but the ratio of variables, ParasoR can locally compute the structure scores such as stem probability or accessibility on demand. A comprehensive analysis of local secondary structures by ParasoR is expected to be a promising way to detect the statistical constraints on long RNAs.

RNA 3D Modeling with FARFAR2, Online.

Understanding the three-dimensional structure of an RNA molecule is often essential to understanding its function. Sampling algorithms and energy functions for RNA structure prediction are improving, due to the increasing diversity of structural data available for training statistical potentials and testing structural data, along with a steady supply of blind challenges through the RNA-Puzzles initiative. The recent FARFAR2 algorithm enables near-native structure predictions on fairly complex RNA structures, including automated selection of final candidate models and estimation of model accuracy. Here, we describe the use of a publicly available webserver for RNA modeling for realistic scenarios using FARFAR2, available at https://rosie.rosettacommons.org/farfar2 . We walk through two cases in some detail: a simple model pseudoknot from the frameshifting element of beet western yellows virus modeled using the "basic interface" to the webserver and a replication of RNA-Puzzle 20, a metagenomic twister sister ribozyme, using the "advanced interface." We also describe example runs of FARFAR2 modeling including two kinds of experimental data: a c-di-GMP riboswitch modeled with low-resolution restraints from MOHCA-seq experiments and a tandem GA motif modeled with 1H NMR chemical shifts.

Examples of Structural Motifs in Viral Genomes and Approaches for RNA Structure Characterization.

The relationship between conserved structural motifs and their biological function in the virus replication cycle is the interest of many researchers around the world. RNA structure is closely related to RNA function. Therefore, technological progress in high-throughput approaches for RNA structure analysis and the development of new ones are very important. In this mini review, we discuss a few perspectives on the structural elements of viral genomes and some methods used for RNA structure prediction and characterization. Based on the recent literature, we describe several examples of studies concerning the viral genomes, especially severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and influenza A virus (IAV). Herein, we emphasize that a better understanding of viral genome architecture allows for the discovery of the structure-function relationship, and as a result, the discovery of new potential antiviral therapeutics.

Learning RNA structure prediction from crowd-designed RNAs.

Learning RNA structure prediction from crowd-designed RNAs.

Determination of 3D structure and molecular interaction for mir-135b and its silencer as Triple Negative Breast Cancer (TNBC) biomarkers

Triple Negative Breast Cancer (TNBC) is considered as the most dangerous breast cancer type in women and the most difficult to be medicated due to its unresponsive nature towards drug treatments. However, as transcriptomics studies are becoming more advanced, the research on non-coding (nc)RNAs like miRNA and siRNA is being considered as a more feasible approach to deal with TNBC. The mir-135b, a newly found miRNA that has a role in the molecular mechanism of TNBC, is studied mainly for its function as a possible biomarker and a drug candidate. The molecular interaction and structures are still unknown and determining them is the objective of this research. This research utilizes the RNAComposer webservice for 3D RNA structure prediction and the AutoDockTool for molecular docking to determine their interaction. The result shows satisfactory illustration of mir-135b and its siRNA structures with feasible bindings between them. It is shown that the bindings are strong enough to maintain silencing of the gene. However, the molecular docking result shows reversible interaction based on the binding energy meaning its biomarker potential requires combination with other special drug delivery systems. It could be inferred that the siRNAs have potential to be developed in the wet laboratory settings.

Functional Interactions of Kluyveromyces lactis Telomerase Reverse Transcriptase with the Three-Way Junction and the Template Domains of Telomerase RNA.

The ribonucleoprotein telomerase contains two essential components: telomerase RNA (TER) and telomerase reverse transcriptase (TERT, Est2 in yeast). A small portion of TER, termed the template, is copied by TERT onto the chromosome ends, thus compensating for sequence loss due to incomplete DNA replication and nuclease action. Although telomerase RNA is highly divergent in sequence and length across fungi and mammals, structural motifs essential for telomerase function are conserved. Here, we show that Est2 from the budding yeast Kluyveromyces lactis (klEst2) binds specifically to an essential three-way junction (TWJ) structure in K. lactis TER, which shares a conserved structure and sequence features with the essential CR4-CR5 domain of vertebrate telomerase RNA. klEst2 also binds specifically to the template domain, independently and mutually exclusive of its interaction with TWJ. Furthermore, we present the high-resolution structure of the klEst2 telomerase RNA-binding domain (klTRBD). Mutations introduced in vivo in klTRBD based on the solved structure or in TWJ based on its predicted RNA structure caused severe telomere shortening. These results demonstrate the conservation and importance of these domains and the multiple protein–RNA interactions between Est2 and TER for telomerase function.

RNAGCN: RNA tertiary structure assessment with a graph convolutional network

RNAs play crucial and versatile roles in cellular biochemical reactions. Since experimental approaches of determining their three-dimensional (3D) structures are costly and less efficient, it is greatly advantageous to develop computational methods to predict RNA 3D structures. For these methods, designing a model or scoring function for structure quality assessment is an essential step but this step poses challenges. In this study, we designed and trained a deep learning model to tackle this problem. The model was based on a graph convolutional network (GCN) and named RNAGCN. The model provided a natural way of representing RNA structures, avoided complex algorithms to preserve atomic rotational equivalence, and was capable of extracting features automatically out of structural patterns. Testing results on two datasets convincingly demonstrated that RNAGCN performs similarly to or better than four leading scoring functions. Our approach provides an alternative way of RNA tertiary structure assessment and may facilitate RNA structure predictions. RNAGCN can be downloaded from https://gitee.com/dcw-RNAGCN/rnagcn.

Linear-Time Algorithms for RNA Structure Prediction.

RNA secondary structure prediction is widely used to understand RNA function. Existing dynamic programming-based algorithms, both the classical minimum free energy (MFE) methods and partition function methods, suffer from a major limitation: their runtimes scale cubically with the RNA length, and this slowness limits their use in genome-wide applications. Inspired by incremental parsing for context-free grammars in computational linguistics, we designed linear-time heuristic algorithms, LinearFold and LinearPartition, to approximate the MFE structure, partition function and base pairing probabilities. These programs are orders of magnitude faster than Vienna RNAfold and CONTRAfold on long sequences. More interestingly, LinearFold and LinearPartition lead to more accurate predictions on the longest sequence families for which the structures are well established (16S and 23S Ribosomal RNAs), as well as improved accuracies for long-range base pairs (500 + nucleotides apart). This chapter provides protocols for using LinearFold and LinearPartition for secondary structure prediction.

Predicting RNA distance-based contact maps by integrated deep learning on physics-inferred secondary structure and evolutionary-derived mutational coupling.

MotivationRecently, AlphaFold2 achieved high experimental accuracy for the majority of proteins in Critical Assessment of Structure Prediction (CASP 14). This raises the hope that one day, we may achieve the same feat for RNA structure prediction for those structured RNAs, which is as fundamentally and practically important similar to protein structure prediction. One major factor in the recent advancement of protein structure prediction is the highly accurate prediction of distance-based contact maps of proteins.ResultsHere, we showed that by integrated deep learning with physics-inferred secondary structures, co-evolutionary information and multiple sequence-alignment sampling, we can achieve RNA contact-map prediction at a level of accuracy similar to that in protein contact-map prediction. More importantly, highly accurate prediction for top L long-range contacts can be assured for those RNAs with a high effective number of homologous sequences (Neff > 50). The initial use of the predicted contact map as distance-based restraints confirmed its usefulness in 3D structure prediction.Availability and implementationSPOT-RNA-2D is available as a web server at https://sparks-lab.org/server/spot-rna-2d/ and as a standalone program at https://github.com/jaswindersingh2/SPOT-RNA-2D.Supplementary information Supplementary data are available at Bioinformatics online.

Secondary structure prediction of long noncoding RNA: review and experimental comparison of existing approaches.

In contrast to messenger RNAs, the function of the wide range of existing long noncoding RNAs (lncRNAs) largely depends on their structure, which determines interactions with partner molecules. Thus, the determination or prediction of the secondary structure of lncRNAs is critical to uncover their function. Classical approaches for predicting RNA secondary structure have been based on dynamic programming and thermodynamic calculations. In the last 4 years, a growing number of machine learning (ML)-based models, including deep learning (DL), have achieved breakthrough performance in structure prediction of biomolecules such as proteins and have outperformed classical methods in short transcripts folding. Nevertheless, the accurate prediction for lncRNA still remains far from being effectively solved. Notably, the myriad of new proposals has not been systematically and experimentally evaluated. In this work, we compare the performance of the classical methods as well as the most recently proposed approaches for secondary structure prediction of RNA sequences using a unified and consistent experimental setup. We use the publicly available structural profiles for 3023 yeast RNA sequences, and a novel benchmark of well-characterized lncRNA structures from different species. Moreover, we propose a novel metric to assess the predictive performance of methods, exclusively based on the chemical probing data commonly used for profiling RNA structures, avoiding any potential bias incorporated by computational predictions when using dot-bracket references. Our results provide a comprehensive comparative assessment of existing methodologies, and a novel and public benchmark resource to aid in the development and comparison of future approaches. Full source code and benchmark datasets are available at: https://github.com/sinc-lab/lncRNA-folding. lbugnon@sinc.unl.edu.ar.

ShapeSorter: a fully probabilistic method for detecting conserved RNAstructure features supported by SHAPE evidence.

There is an increased interest in the determination of RNA structures in vivo as it is now possible to probe them in a high-throughput manner, e.g. using SHAPE protocols. By now, there exist a range of computational methods that integrate experimental SHAPE-probing evidence into computational RNA secondary structure prediction. The state-of-the-art in this field is currently provided by computational methods that employ the minimum-free energy strategy for prediction RNA secondary structures with SHAPE-probing evidence. These methods, however, rely on the assumption that transcripts in vivo fold into the thermodynamically most stable configuration and ignore evolutionary evidence for conserved RNA structure features. We here present a new computational method, ShapeSorter, that predicts RNA structure features without employing the thermodynamic strategy. Instead, ShapeSorter employs a fully probabilistic framework to identify RNA structure features that are supported by evolutionary and SHAPE-probing evidence. Our method can capture RNA structure heterogeneity, pseudo-knotted RNA structures as well as transient and mutually exclusive RNA structure features. Moreover, it estimates P-values for the predicted RNA structure features which allows for easy filtering and ranking. We investigate the merits of our method in a comprehensive performance benchmarking and conclude that ShapeSorter has a significantly superior performance for predicting base-pairs than the existing state-of-the-art methods.

Dual role of CsrA in regulating the hemolytic activity of Escherichia coli O157:H7

ABSTRACT Post-transcriptional global carbon storage regulator A (CsrA) is a sequence-specific RNA-binding protein involved in the regulation of multiple bacterial processes. Hemolysin is an important virulence factor in the enterohemorrhagic Escherichia coli O157:H7 (EHEC). Here, we show that CsrA plays a dual role in the regulation of hemolysis in EHEC. CsrA significantly represses plasmid-borne enterohemolysin (EhxA)-mediated hemolysis and activates chromosome-borne hemolysin E (HlyE)-mediated hemolysis through different mechanisms. RNA structure prediction revealed a well-matched stem-loop structure with two potential CsrA binding sites located on the 5' untranslated region (UTR) of ehxB, which encodes a translocator required for EhxA secretion. CsrA inhibits EhxA secretion by directly binding to the RNA leader sequence of ehxB to repress its expression in two different ways: CsrA either binds to the Shine–Dalgarno sequence of ehxB to block ribosome access or to ehxB transcript to promote its mRNA decay. The predicted CsrA-binding site 1 of ehxB is essential for its regulation. There is a single potential CsrA-binding site at the 5'-end of the hlyE transcript, and its mutation completely abolishes CsrA-dependent activation. CsrA can also stabilize hlyE mRNA by directly binding to its 5' UTR. Overall, our results indicate that CsrA acts as a hemolysis modulator to regulate pathogenicity under certain conditions.

RNA structure prediction with direct coupling analysis

Predicting the structure of biological molecules such as proteins and RNA has become an exciting challenge in recent years. Structure determination using experimental methods is laborious, so alternative approaches using the increasingly powerful HPC systems is promising.