Unifying evolutionary and thermodynamic information for RNA folding of multiple alignments

Comparative analysis of RNA 3D structure prediction methods: towards enhanced modeling of RNA-ligand interactions

Accurate RNA structure models are crucial for designing small molecule ligands that modulate their functions. This study assesses six standalone RNA 3D structure prediction methods-DeepFoldRNA, RhoFold, BRiQ, FARFAR2, SimRNA and Vfold2, excluding web-based tools due to intellectual property concerns. We focus on reproducing the RNA structure existing in RNA-small molecule complexes, particularly on the ability to model ligand binding sites. Using a comprehensive set of RNA structures from the PDB, which includes diverse structural elements, we found that machine learning (ML)-based methods effectively predict global RNA folds but are less accurate with local interactions. Conversely, non-ML-based methods demonstrate higher precision in modeling intramolecular interactions, particularly with secondary structure restraints. Importantly, ligand-binding site accuracy can remain sufficiently high for practical use, even if the overall model quality is not optimal. With the recent release of AlphaFold 3, we included this advanced method in our tests. Benchmark subsets containing new structures, not used in the training of the tested ML methods, show that AlphaFold 3's performance was comparable to other ML-based methods, albeit with some challenges in accurately modeling ligand binding sites. This study underscores the importance of enhancing binding site prediction accuracy and the challenges in modeling RNA-ligand interactions accurately.

Phylogenetic and Chemical Probing Information as Soft Constraints in RNA Secondary Structure Prediction

Extrinsic, experimental information can be incorporated into thermodynamics-based RNA folding algorithms in the form of pseudo-energies. Evolutionary conservation of RNA secondary structure elements is detectable in alignments of phylogenetically related sequences and provides evidence for the presence of certain base pairs that can also be converted into pseudo-energy contributions. We show that the centroid base pairs computed from a consensus folding model such as RNAalifold result in a substantial improvement of the prediction accuracy for single sequences. Evidence for specific base pairs turns out to be more informative than a position-wise profile for the conservation of the pairing status. A comparison with chemical probing data, furthermore, strongly suggests that phylogenetic base pairing data are more informative than position-specific data on (un)pairedness as obtained from chemical probing experiments. In this context we demonstrate, in addition, that the conversion of signal from probing data into pseudo-energies is possible using thermodynamic structure predictions as a reference instead of known RNA structures.

Comparative RNA Genomics

Over the last quarter of a century it has become clear that RNA is much more than just a boring intermediate in protein expression. Ancient RNAs still appear in the core information metabolism and comprise a surprisingly large component in bacterial gene regulation. A common theme with these types of mostly small RNAs is their reliance of conserved secondary structures. Large-scale sequencing projects, on the other hand, have profoundly changed our understanding of eukaryotic genomes. Pervasively transcribed, they give rise to a plethora of large and evolutionarily extremely flexible non-coding RNAs that exert a vastly diverse array of molecule functions. In this chapter we provide a-necessarily incomplete-overview of the current state of comparative analysis of non-coding RNAs, emphasizing computational approaches as a means to gain a global picture of the modern RNA world.

Evolutionary Structure Conservation and Covariance Scores

Effective homology search for non-coding RNAs is frequently not possible via sequence similarity alone. Current methods leverage evolutionary information like structure conservation or covariance scores to identify homologs in organisms that are phylogenetically more distant. In this chapter, we introduce the theoretical background of evolutionary structure conservation and covariance score, and we show hands-on how current methods in the field are applied on example datasets.

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

RNAs are fundamental in living cells and perform critical functions determined by their tertiary architectures. However, accurate modeling of 3D RNA structure remains a challenging problem. 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 RNA structure assembly. The method significantly outperforms previous approaches by >73.3% in TM-score on a sequence-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 with fast training protocol allows large-scale application of high-resolution RNA structure modeling and can be further improved with future expansion of RNA structure databases.

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.

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.

The Bacillaceae-1 RNA motif comprises two distinct classes

Non-coding RNAs are key regulatory players in bacteria. Many computationally predicted non-coding RNAs, however, lack functional associations. An example is the Bacillaceae-1 RNA motif, whose Rfam model consists of two hairpin loops. We find the motif conserved in nine of 13 non-pathogenic strains of the genus Bacillus but only in one pathogenic strain. To elucidate functional characteristics, we studied 118 hits of the Rfam model in 11 Bacillus spp. and found two distinct classes based on the ensemble diversity of their RNA secondary structure and the genomic context concerning the ribosomal RNA (rRNA) cluster. Forty hits are associated with the rRNA cluster, of which all 19 hits upstream flanking of 16S rRNA have a reverse complementary structure of low structural diversity. Fifty-two hits have large ensemble diversity, of which 38 are located between two coding genes. For eight hits in Bacillus subtilis, we investigated public expression data under various conditions and observed either the forward or the reverse complementary motif expressed. Five hits are associated with the rRNA cluster. Four of them are located upstream of the 16S rRNA and are not transcriptionally active, but instead, their reverse complements with low structural diversity are expressed together with the rRNA cluster. The three other hits are located between two coding genes in non-conserved genomic loci. Two of them are independently expressed from their surrounding genes and are structurally diverse. In summary, we found that Bacillaceae-1 RNA motifs upstream flanking of ribosomal RNA clusters tend to have one stable structure with the reverse complementary motif expressed in B. subtilis. In contrast, a subgroup of intergenic motifs has the thermodynamic potential for structural switches.

ConsAlifold: considering RNA structural alignments improves prediction accuracy of RNA consensus secondary structures

MOTIVATION

By detecting homology among RNAs, the probabilistic consideration of RNA structural alignments has improved the prediction accuracy of significant RNA prediction problems. Predicting an RNA consensus secondary structure from an RNA sequence alignment is a fundamental research objective because in the detection of conserved base-pairings among RNA homologs, predicting an RNA consensus secondary structure is more convenient than predicting an RNA structural alignment.

RESULTS

We developed and implemented ConsAlifold, a dynamic programming-based method that predicts the consensus secondary structure of an RNA sequence alignment. ConsAlifold considers RNA structural alignments. ConsAlifold achieves moderate running time and the best prediction accuracy of RNA consensus secondary structures among available prediction methods.

AVAILABILITY AND IMPLEMENTATION

ConsAlifold, data and Python scripts for generating both figures and tables are freely available at https://github.com/heartsh/consalifold.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

RNA inter-nucleotide 3D closeness prediction by deep residual neural networks

MOTIVATION

Recent years have witnessed that the inter-residue contact/distance in proteins could be accurately predicted by deep neural networks, which significantly improve the accuracy of predicted protein structure models. In contrast, fewer studies have been done for the prediction of RNA inter-nucleotide 3D closeness.

RESULTS

We proposed a new algorithm named RNAcontact for the prediction of RNA inter-nucleotide 3D closeness. RNAcontact was built based on the deep residual neural networks. The covariance information from multiple sequence alignments and the predicted secondary structure were used as the input features of the networks. Experiments show that RNAcontact achieves the respective precisions of 0.8 and 0.6 for the top L/10 and L (where L is the length of an RNA) predictions on an independent test set, significantly higher than other evolutionary coupling methods. Analysis shows that about 1/3 of the correctly predicted 3D closenesses are not base pairings of secondary structure, which are critical to the determination of RNA structure. In addition, we demonstrated that the predicted 3D closeness could be used as distance restraints to guide RNA structure folding by the 3dRNA package. More accurate models could be built by using the predicted 3D closeness than the models without using 3D closeness.

AVAILABILITY AND IMPLEMENTATION

The webserver and a standalone package are available at: http://yanglab.nankai.edu.cn/RNAcontact/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Modeling of Three-Dimensional RNA Structures Using SimRNA

The molecules of the ribonucleic acid (RNA) perform a variety of vital roles in all living cells. Their biological function depends on their structure and dynamics, both of which are difficult to experimentally determine but can be theoretically inferred based on the RNA sequence. SimRNA is one of the computational methods for molecular simulations of RNA 3D structure formation. The method is based on a simplified (coarse-grained) representation of nucleotide chains, a statistically derived model of interactions (statistical potential), and the Monte Carlo method as a conformational sampling scheme.The current version of SimRNA (3.22) is able to predict basic topologies of RNA molecules with sizes up to about 50-70 nucleotides, based on their sequences only, and larger molecules if supplied with appropriate distance restraints. The user can specify various types of restraints, including secondary structure, pairwise atom-atom distances, and positions of atoms. SimRNA can be also used for studying systems composed of several chains of RNA. SimRNA is a folding simulations method, thus it allows for examining folding pathways, getting an approximate view of the energy landscapes.

RNA 3D structure prediction guided by independent folding of homologous sequences

BACKGROUND

The understanding of the importance of RNA has dramatically changed over recent years. As in the case of proteins, the function of an RNA molecule is encoded in its tertiary structure, which in turn is determined by the molecule's sequence. The prediction of tertiary structures of complex RNAs is still a challenging task.

RESULTS

Using the observation that RNA sequences from the same RNA family fold into conserved structure, we test herein whether parallel modeling of RNA homologs can improve ab initio RNA structure prediction. EvoClustRNA is a multi-step modeling process, in which homologous sequences for the target sequence are selected using the Rfam database. Subsequently, independent folding simulations using Rosetta FARFAR and SimRNA are carried out. The model of the target sequence is selected based on the most common structural arrangement of the common helical fragments. As a test, on two blind RNA-Puzzles challenges, EvoClustRNA predictions ranked as the first of all submissions for the L-glutamine riboswitch and as the second for the ZMP riboswitch. Moreover, through a benchmark of known structures, we discovered several cases in which particular homologs were unusually amenable to structure recovery in folding simulations compared to the single original target sequence.

CONCLUSION

This work, for the first time to our knowledge, demonstrates the importance of the selection of the target sequence from an alignment of an RNA family for the success of RNA 3D structure prediction. These observations prompt investigations into a new direction of research for checking 3D structure "foldability" or "predictability" of related RNA sequences to obtain accurate predictions. To support new research in this area, we provide all relevant scripts in a documented and ready-to-use form. By exploring new ideas and identifying limitations of the current RNA 3D structure prediction methods, this work is bringing us closer to the near-native computational RNA 3D models.

Translational co-regulation of a ligand and inhibitor by a conserved RNA element

In many organisms, transcriptional and post-transcriptional regulation of components of pathways or processes has been reported. However, to date, there are few reports of translational co-regulation of multiple components of a developmental signaling pathway. Here, we show that an RNA element which we previously identified as a dorsal localization element (DLE) in the 3'UTR of zebrafish nodal-related1/squint (ndr1/sqt) ligand mRNA, is shared by the related ligand nodal-related2/cyclops (ndr2/cyc) and the nodal inhibitors, lefty1 (lft1) and lefty2 mRNAs. We investigated the activity of the DLEs through functional assays in live zebrafish embryos. The lft1 DLE localizes fluorescently labeled RNA similarly to the ndr1/sqt DLE. Similar to the ndr1/sqt 3'UTR, the lft1 and lft2 3'UTRs are bound by the RNA-binding protein (RBP) and translational repressor, Y-box binding protein 1 (Ybx1), whereas deletions in the DLE abolish binding to Ybx1. Analysis of zebrafish ybx1 mutants shows that Ybx1 represses lefty1 translation in embryos. CRISPR/Cas9-mediated inactivation of human YBX1 also results in human NODAL translational de-repression, suggesting broader conservation of the DLE RNA element/Ybx1 RBP module in regulation of Nodal signaling. Our findings demonstrate translational co-regulation of components of a signaling pathway by an RNA element conserved in both sequence and structure and an RBP, revealing a 'translational regulon'.

Structure and Interaction Prediction in Prokaryotic RNA Biology

Many years of research in RNA biology have soundly established the importance of RNA-based regulation far beyond most early traditional presumptions. Importantly, the advances in "wet" laboratory techniques have produced unprecedented amounts of data that require efficient and precise computational analysis schemes and algorithms. Hence, many in silico methods that attempt topological and functional classification of novel putative RNA-based regulators are available. In this review, we technically outline thermodynamics-based standard RNA secondary structure and RNA-RNA interaction prediction approaches that have proven valuable to the RNA research community in the past and present. For these, we highlight their usability with a special focus on prokaryotic organisms and also briefly mention recent advances in whole-genome interactomics and how this may influence the field of predictive RNA research.

Multiple Sequence Alignments Enhance Boundary Definition of RNA Structures

Self-contained structured domains of RNA sequences have often distinct molecular functions. Determining the boundaries of structured domains of a non-coding RNA (ncRNA) is needed for many ncRNA gene finder programs that predict RNA secondary structures in aligned genomes because these methods do not necessarily provide precise information about the boundaries or the location of the RNA structure inside the predicted ncRNA. Even without having a structure prediction, it is of interest to search for structured domains, such as for finding common RNA motifs in RNA-protein binding assays. The precise definition of the boundaries are essential for downstream analyses such as RNA structure modelling, e.g., through covariance models, and RNA structure clustering for the search of common motifs. Such efforts have so far been focused on single sequences, thus here we present a comparison for boundary definition between single sequence and multiple sequence alignments. We also present a novel approach, named RNAbound, for finding the boundaries that are based on probabilities of evolutionarily conserved base pairings. We tested the performance of two different methods on a limited number of Rfam families using the annotated structured RNA regions in the human genome and their multiple sequence alignments created from 14 species. The results show that multiple sequence alignments improve the boundary prediction for branched structures compared to single sequences independent of the chosen method. The actual performance of the two methods differs on single hairpin structures and branched structures. For the RNA families with branched structures, including transfer RNA (tRNA) and small nucleolar RNAs (snoRNAs), RNAbound improves the boundary predictions using multiple sequence alignments to median differences of −6 and −11.5 nucleotides (nts) for left and right boundary, respectively (window size of 200 nts).

Comparative RNA Genomics

Over the last two decades it has become clear that RNA is much more than just a boring intermediate in protein expression. Ancient RNAs still appear in the core information metabolism and comprise a surprisingly large component in bacterial gene regulation. A common theme with these types of mostly small RNAs is their reliance of conserved secondary structures. Large scale sequencing projects, on the other hand, have profoundly changed our understanding of eukaryotic genomes. Pervasively transcribed, they give rise to a plethora of large and evolutionarily extremely flexible noncoding RNAs that exert a vastly diverse array of molecule functions. In this chapter we provide a-necessarily incomplete-overview of the current state of comparative analysis of noncoding RNAs, emphasizing computational approaches as a means to gain a global picture of the modern RNA world.

Identification of the functional alteration signatures across different cancer types with support vector machine and feature analysis

Cancers are regarded as malignant proliferations of tumor cells present in many tissues and organs, which can severely curtail the quality of human life. The potential of using plasma DNA for cancer detection has been widely recognized, leading to the need of mapping the tissue-of-origin through the identification of somatic mutations. With cutting-edge technologies, such as next-generation sequencing, numerous somatic mutations have been identified, and the mutation signatures have been uncovered across different cancer types. However, somatic mutations are not independent events in carcinogenesis but exert functional effects. In this study, we applied a pan-cancer analysis to five types of cancers: (I) breast cancer (BRCA), (II) colorectal adenocarcinoma (COADREAD), (III) head and neck squamous cell carcinoma (HNSC), (IV) kidney renal clear cell carcinoma (KIRC), and (V) ovarian cancer (OV). Based on the mutated genes of patients suffering from one of the aforementioned cancer types, patients they were encoded into a large number of numerical values based upon the enrichment theory of gene ontology (GO) terms and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. We analyzed these features with the Monte-Carlo Feature Selection (MCFS) method, followed by the incremental feature selection (IFS) method to identify functional alteration features that could be used to build the support vector machine (SVM)-based classifier for distinguishing the five types of cancers. Our results showed that the optimal classifier with the selected 344 features had the highest Matthews correlation coefficient value of 0.523. Sixteen decision rules produced by the MCFS method can yield an overall accuracy of 0.498 for the classification of the five cancer types. Further analysis indicated that some of these features and rules were supported by previous experiments. This study not only presents a new approach to mapping the tissue-of-origin for cancer detection but also unveils the specific functional alterations of each cancer type, providing insight into cancer-specific functional aberrations as potential therapeutic targets. This article is part of a Special Issue entitled: Accelerating Precision Medicine through Genetic and Genomic Big Data Analysis edited by Yudong Cai & Tao Huang.

Alignment-free comparative genomic screen for structured RNAs using coarse-grained secondary structure dot plots

Background

Structured non-coding RNAs play many different roles in the cells, but the annotation of these RNAs is lacking even within the human genome. The currently available computational tools are either too computationally heavy for use in full genomic screens or rely on pre-aligned sequences.

Methods

Here we present a fast and efficient method, DotcodeR, for detecting structurally similar RNAs in genomic sequences by comparing their corresponding coarse-grained secondary structure dot plots at string level. This allows us to perform an all-against-all scan of all window pairs from two genomes without alignment.

Results

Our computational experiments with simulated data and real chromosomes demonstrate that the presented method has good sensitivity.

Conclusions

DotcodeR can be useful as a pre-filter in a genomic comparative scan for structured RNAs.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-017-4309-y) contains supplementary material, which is available to authorized users.

In silico approaches to RNA aptamer design

RNA aptamers are ribonucleic acids that bind to specific target molecules. An RNA aptamer for a disease-related protein has great potential for development into a new drug. However, huge time and cost investments are required to develop an RNA aptamer into a pharmaceutical. Recently, SELEX combined with high-throughput sequencers (i.e., HT-SELEX) has been widely used to select candidate RNA aptamers that bind to a target protein with high affinity and specificity. After candidate selection, further optimizations such as shortening and modifying candidate sequences are performed. In these steps, in silico approaches are expected to reduce the time and cost associated with aptamer drug development. In this article, we review existing in silico approaches to RNA aptamer development, including a method for ranking the candidates of RNA aptamers from HT-SELEX data, clustering a huge number of aptamer sequences, and finding motifs amidst a set of significant RNA aptamers. It is expected that further studies in addition to these methods will be utilized for in silico RNA aptamer design, permitting a minimal number of experiments to be performed through the utilization of sophisticated computational methods.

RNAscClust: clustering RNA sequences using structure conservation and graph based motifs

MOTIVATION

Clustering RNA sequences with common secondary structure is an essential step towards studying RNA function. Whereas structural RNA alignment strategies typically identify common structure for orthologous structured RNAs, clustering seeks to group paralogous RNAs based on structural similarities. However, existing approaches for clustering paralogous RNAs, do not take the compensatory base pair changes obtained from structure conservation in orthologous sequences into account.

RESULTS

Here, we present RNAscClust , the implementation of a new algorithm to cluster a set of structured RNAs taking their respective structural conservation into account. For a set of multiple structural alignments of RNA sequences, each containing a paralog sequence included in a structural alignment of its orthologs, RNAscClust computes minimum free-energy structures for each sequence using conserved base pairs as prior information for the folding. The paralogs are then clustered using a graph kernel-based strategy, which identifies common structural features. We show that the clustering accuracy clearly benefits from an increasing degree of compensatory base pair changes in the alignments.

AVAILABILITY AND IMPLEMENTATION

RNAscClust is available at http://www.bioinf.uni-freiburg.de/Software/RNAscClust .

CONTACT

gorodkin@rth.dk or backofen@informatik.uni-freiburg.de.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Predicting RNA secondary structures from sequence and probing data

RNA secondary structures have proven essential for understanding the regulatory functions performed by RNA such as microRNAs, bacterial small RNAs, or riboswitches. This success is in part due to the availability of efficient computational methods for predicting RNA secondary structures. Recent advances focus on dealing with the inherent uncertainty of prediction by considering the ensemble of possible structures rather than the single most stable one. Moreover, the advent of high-throughput structural probing has spurred the development of computational methods that incorporate such experimental data as auxiliary information.

Full-length RNA structure prediction of the HIV-1 genome reveals a conserved core domain

A distance constrained secondary structural model of the ≈10 kb RNA genome of the HIV-1 has been predicted but higher-order structures, involving long distance interactions, are currently unknown. We present the first global RNA secondary structure model for the HIV-1 genome, which integrates both comparative structure analysis and information from experimental data in a full-length prediction without distance constraints. Besides recovering known structural elements, we predict several novel structural elements that are conserved in HIV-1 evolution. Our results also indicate that the structure of the HIV-1 genome is highly variable in most regions, with a limited number of stable and conserved RNA secondary structures. Most interesting, a set of long distance interactions form a core organizing structure (COS) that organize the genome into three major structural domains. Despite overlapping protein-coding regions the COS is supported by a particular high frequency of compensatory base changes, suggesting functional importance for this element. This new structural element potentially organizes the whole genome into three major domains protruding from a conserved core structure with potential roles in replication and evolution for the virus.

Finding consensus stable local optimal structures for aligned RNA sequences and its application to discovering riboswitch elements

Many non-coding RNAs (ncRNAs) can fold into alternate native structures and perform different biological functions. The computational prediction of an ncRNA's alternate native structures can be conducted by analysing the ncRNA's energy landscape. Previously, we have developed a computational approach, RNASLOpt, to predict alternate native structures for a single RNA. In this paper, in order to improve the accuracy of the prediction, we incorporate structural conservation information among a family of related ncRNA sequences to the prediction. We propose a comparative approach, RNAConSLOpt, to produce all possible consensus SLOpt stack configurations that are conserved on the consensus energy landscape of a family of related ncRNAs. Benchmarking tests show that RNAConSLOpt can reduce the number of candidate structures compared with RNASLOpt, and can predict ncRNAs' alternate native structures accurately. Moreover, an application of the proposed pipeline to bacteria in Bacillus genus has discovered several novel riboswitch candidates.

RNA secondary structure prediction from multi-aligned sequences

It has been well accepted that the RNA secondary structures of most functional non-coding RNAs (ncRNAs) are closely related to their functions and are conserved during evolution. Hence, prediction of conserved secondary structures from evolutionarily related sequences is one important task in RNA bioinformatics; the methods are useful not only to further functional analyses of ncRNAs but also to improve the accuracy of secondary structure predictions and to find novel functional RNAs from the genome. In this review, I focus on common secondary structure prediction from a given aligned RNA sequence, in which one secondary structure whose length is equal to that of the input alignment is predicted. I systematically review and classify existing tools and algorithms for the problem, by utilizing the information employed in the tools and by adopting a unified viewpoint based on maximum expected gain (MEG) estimators. I believe that this classification will allow a deeper understanding of each tool and provide users with useful information for selecting tools for common secondary structure predictions.

Computational modeling of RNA 3D structures, with the aid of experimental restraints

In addition to mRNAs whose primary function is transmission of genetic information from DNA to proteins, numerous other classes of RNA molecules exist, which are involved in a variety of functions, such as catalyzing biochemical reactions or performing regulatory roles. In analogy to proteins, the function of RNAs depends on their structure and dynamics, which are largely determined by the ribonucleotide sequence. Experimental determination of high-resolution RNA structures is both laborious and difficult, and therefore, the majority of known RNAs remain structurally uncharacterized. To address this problem, computational structure prediction methods were developed that simulate either the physical process of RNA structure formation ("Greek science" approach) or utilize information derived from known structures of other RNA molecules ("Babylonian science" approach). All computational methods suffer from various limitations that make them generally unreliable for structure prediction of long RNA sequences. However, in many cases, the limitations of computational and experimental methods can be overcome by combining these two complementary approaches with each other. In this work, we review computational approaches for RNA structure prediction, with emphasis on implementations (particular programs) that can utilize restraints derived from experimental analyses. We also list experimental approaches, whose results can be relatively easily used by computational methods. Finally, we describe case studies where computational and experimental analyses were successfully combined to determine RNA structures that would remain out of reach for each of these approaches applied separately.

Bioinformatics of prokaryotic RNAs

The genome of most prokaryotes gives rise to surprisingly complex transcriptomes, comprising not only protein-coding mRNAs, often organized as operons, but also harbors dozens or even hundreds of highly structured small regulatory RNAs and unexpectedly large levels of anti-sense transcripts. Comprehensive surveys of prokaryotic transcriptomes and the need to characterize also their non-coding components is heavily dependent on computational methods and workflows, many of which have been developed or at least adapted specifically for the use with bacterial and archaeal data. This review provides an overview on the state-of-the-art of RNA bioinformatics focusing on applications to prokaryotes.

Computational prediction of RNA-RNA interactions

We describe different tools and approaches for RNA-RNA interaction prediction. Recognition of ncRNA targets is predominantly governed by two principles, namely the stability of the duplex between the two interacting RNAs and the internal structure of both mRNA and ncRNA. Thus, approaches can be distinguished into different major categories depending on how they consider inter- and intramolecular structure. The first class completely neglects the internal structure and measures only the stability of the duplex. The second class of approaches abstracts from specific intramolecular structures and uses an ensemble-based approach to calculate the effect of internal structure on a putative binding site, thus measuring the accessibility of the binding sites.Since accessibility-based approaches can handle only one continuous interaction site, two addition types of approaches were introduced which predict a joint structure for the interacting RNAs. Since this problem is NP-complete, the approaches can handle only a restricted class of joint structures. The first are co-folding approaches, which predict a joint structure that is nested when the both sequences are concatenated. The last and most complex class of approaches impose only the restriction that they discard zipper-like structures. Finally, we will discuss the use of conservation information in RNA-target prediction.

Energy-based RNA consensus secondary structure prediction in multiple sequence alignments

Many biologically important RNA structures are conserved in evolution leading to characteristic mutational patterns. RNAalifold is a widely used program to predict consensus secondary structures in multiple alignments by combining evolutionary information with traditional energy-based RNA folding algorithms. Here we describe the theory and applications of the RNAalifold algorithm. Consensus secondary structure prediction not only leads to significantly more accurate structure models, but it also allows to study structural conservation of functional RNAs.

RNAsnp: efficient detection of local RNA secondary structure changes induced by SNPs

Structural characteristics are essential for the functioning of many noncoding RNAs and cis-regulatory elements of mRNAs. SNPs may disrupt these structures, interfere with their molecular function, and hence cause a phenotypic effect. RNA folding algorithms can provide detailed insights into structural effects of SNPs. The global measures employed so far suffer from limited accuracy of folding programs on large RNAs and are computationally too demanding for genome-wide applications. Here, we present a strategy that focuses on the local regions of maximal structural change between mutant and wild-type. These local regions are approximated in a “screening mode” that is intended for genome-wide applications. Furthermore, localized regions are identified as those with maximal discrepancy. The mutation effects are quantified in terms of empirical P values. To this end, the RNAsnp software uses extensive precomputed tables of the distribution of SNP effects as function of length and GC content. RNAsnp thus achieves both a noise reduction and speed-up of several orders of magnitude over shuffling-based approaches. On a data set comprising 501 SNPs associated with human-inherited diseases, we predict 54 to have significant local structural effect in the untranslated region of mRNAs. RNAsnp is available at http://rth.dk/resources/rnasnp.

Computational detection of abundant long-range nucleotide covariation in Drosophila genomes

Functionally important nucleotide base-pairing often manifests itself in sequence alignments in the form of compensatory base changes (covariation). We developed a novel index-based computational method (CovaRNA) to detect long-range covariation on a genomic scale, as well as another computational method (CovStat) for determining the statistical significance of observed covariation patterns in alignment pairs. Here we present an all-versus-all search for nucleotide covariation in Drosophila genomic alignments. The search is genome wide, with the restriction that only alignments that correspond to euchromatic regions, which consist of at least 10 Drosophila species, are being considered (59% of the euchromatic genome of Drosophila melanogaster). We find that long-range covariations are especially prevalent between exons of mRNAs as well as noncoding RNAs; the majority of the observed covariations appear as not reverse complementary, but as synchronized mutations, which could be due to interactions with common interaction partners or due to the involvement of genomic elements that are antisense of annotated transcripts. The involved genes are enriched for functions related to regionalization as well as neural and developmental processes. These results are computational evidence that RNA-RNA long-range interactions are a widespread phenomenon that is of fundamental importance to a variety of cellular processes.

MicroRNA discovery by similarity search to a database of RNA-seq profiles

In silico generated search for microRNAs (miRNAs) has been driven by methods compiling structural features of the miRNA precursor hairpin, as well as to some degree combining this with the analysis of RNA-seq profiles for which the miRNA typically leave the drosha/dicer fingerprint of 1-2 ~22 nt blocks of reads corresponding to the mature and star miRNA. In complement to the previous methods, we present a study where we systematically exploit these patterns of read profiles. We created two datasets comprised of 2540 and 4795 read profiles obtained after preprocessing short RNA-seq data from miRBase and ENCODE, respectively. Out of 4795 ENCODE read profiles, 1361 are annotated as non-coding RNAs (ncRNAs) and of which 285 are further annotated as miRNAs. Using deepBlockAlign (dba), we align ncRNA read profiles from ENCODE against the miRBase read profiles (cleaned for "self-matches") and are able to separate ENCODE miRNAs from the other ncRNAs by a Matthews Correlation Coefficient (MCC) of 0.8 and obtain an area under the curve of 0.93. Based on the dba score cut-off of 0.7 at which we observed the maximum MCC of 0.8, we predict 523 novel miRNA candidates. An additional RNA secondary structure analysis reveal that 42 of the candidates overlap with predicted conserved secondary structure. Further analysis reveal that the 523 miRNA candidates are located in genomic regions with MAF block (UCSC) fragmentation and poor sequence conservation, which in part might explain why they have been overlooked in previous efforts. We further analyzed known human and mouse miRNA read profiles and found two distinct classes; the first containing two blocks and the second containing >2 blocks of reads. Also the latter class holds read profiles that have less well defined arrangement of reads in comparison to the former class. On comparison of miRNA read profiles from plants and animals, we observed kingdom specific read profiles that are distinct in terms of both length and distribution of reads within the read profiles to each other. All the data, as well as a server to search miRBase read profiles by uploading a BED file, is available at http://rth.dk/resources/mirdba.

Direct updating of an RNA base-pairing probability matrix with marginal probability constraints

A base-pairing probability matrix (BPPM) stores the probabilities for every possible base pair in an RNA sequence and has been used in many algorithms in RNA informatics (e.g., RNA secondary structure prediction and motif search). In this study, we propose a novel algorithm to perform iterative updates of a given BPPM, satisfying marginal probability constraints that are (approximately) given by recently developed biochemical experiments, such as SHAPE, PAR, and FragSeq. The method is easily implemented and is applicable to common models for RNA secondary structures, such as energy-based or machine-learning-based models. In this article, we focus mainly on the details of the algorithms, although preliminary computational experiments will also be presented.

Quantifying variances in comparative RNA secondary structure prediction

Background

With the advancement of next-generation sequencing and transcriptomics technologies, regulatory effects involving RNA, in particular RNA structural changes are being detected. These results often rely on RNA secondary structure predictions. However, current approaches to RNA secondary structure modelling produce predictions with a high variance in predictive accuracy, and we have little quantifiable knowledge about the reasons for these variances.

Results

In this paper we explore a number of factors which can contribute to poor RNA secondary structure prediction quality. We establish a quantified relationship between alignment quality and loss of accuracy. Furthermore, we define two new measures to quantify uncertainty in alignment-based structure predictions. One of the measures improves on the “reliability score” reported by PPfold, and considers alignment uncertainty as well as base-pair probabilities. The other measure considers the information entropy for SCFGs over a space of input alignments.

Conclusions

Our predictive accuracy improves on the PPfold reliability score. We can successfully characterize many of the underlying reasons for and variances in poor prediction. However, there is still variability unaccounted for, which we therefore suggest comes from the RNA secondary structure predictive model itself.

Incorporating phylogenetic-based covarying mutations into RNAalifold for RNA consensus structure prediction

Background

RNAalifold, a popular computational method for RNA consensus structure prediction, incorporates covarying mutations into a thermodynamic model to fold the aligned RNA sequences. When quantifying covariance, it evaluates conserved signals of two aligned columns with base-pairing rules. This scoring scheme performs better than some other approaches, such as mutual information. However it ignores the phylogenetic history of the aligned sequences, which is an important criterion to evaluate the level of sequence covariance.

Results

In this article, in order to improve the accuracy of consensus structure folding, we propose a novel approach named PhyloRNAalifold. It incorporates the number of covarying mutations on the phylogenetic tree of the aligned sequences into the covariance scoring of RNAalifold. The benchmarking results show that the new scoring scheme of PhyloRNAalifold can improve the consensus structure detection of RNAalifold.

Conclusion

Incorporating additional phylogenetic information of aligned sequences into the covariance scoring of RNAalifold can improve its performance of consensus structures folding. This improvement is correlated with alignment characteristics, such as pair-wise identity and the number of sequences in the alignment.

Effects of using coding potential, sequence conservation and mRNA structure conservation for predicting pyrrolysine containing genes

BACKGROUND

Pyrrolysine (the 22nd amino acid) is in certain organisms and under certain circumstances encoded by the amber stop codon, UAG. The circumstances driving pyrrolysine translation are not well understood. The involvement of a predicted mRNA structure in the region downstream UAG has been suggested, but the structure does not seem to be present in all pyrrolysine incorporating genes.

RESULTS

We propose a strategy to predict pyrrolysine encoding genes in genomes of archaea and bacteria. We cluster open reading frames interrupted by the amber codon based on sequence similarity. We rank these clusters according to several features that may influence pyrrolysine translation. The ranking effects of different features are assessed and we propose a weighted combination of these features which best explains the currently known pyrrolysine incorporating genes. We devote special attention to the effect of structural conservation and provide further substantiation to support that structural conservation may be influential - but is not a necessary factor. Finally, from the weighted ranking, we identify a number of potentially pyrrolysine incorporating genes.

CONCLUSIONS

We propose a method for prediction of pyrrolysine incorporating genes in genomes of bacteria and archaea leading to insights about the factors driving pyrrolysine translation and identification of new gene candidates. The method predicts known conserved genes with high recall and predicts several other promising candidates for experimental verification. The method is implemented as a computational pipeline which is available on request.

GraphClust: alignment-free structural clustering of local RNA secondary structures

Motivation: Clustering according to sequence–structure similarity has now become a generally accepted scheme for ncRNA annotation. Its application to complete genomic sequences as well as whole transcriptomes is therefore desirable but hindered by extremely high computational costs.

Results: We present a novel linear-time, alignment-free method for comparing and clustering RNAs according to sequence and structure. The approach scales to datasets of hundreds of thousands of sequences. The quality of the retrieved clusters has been benchmarked against known ncRNA datasets and is comparable to state-of-the-art sequence–structure methods although achieving speedups of several orders of magnitude. A selection of applications aiming at the detection of novel structural ncRNAs are presented. Exemplarily, we predicted local structural elements specific to lincRNAs likely functionally associating involved transcripts to vital processes of the human nervous system. In total, we predicted 349 local structural RNA elements.

Availability: The GraphClust pipeline is available on request.

Contact:backofen@informatik.uni-freiburg.de

Supplementary information:Supplementary data are available at Bioinformatics online.

Discovery of a divergent HPIV4 from respiratory secretions using second and third generation metagenomic sequencing

Molecular detection of viruses has been aided by high-throughput sequencing, permitting the genomic characterization of emerging strains. In this study, we comprehensively screened 500 respiratory secretions from children with upper and/or lower respiratory tract infections for viral pathogens. The viruses detected are described, including a divergent human parainfluenza virus type 4 from GS FLX pyrosequencing of 92 specimens. Complete full-genome characterization of the virus followed, using Single Molecule, Real-Time (SMRT) sequencing. Subsequent "primer walking" combined with Sanger sequencing validated the RS platform's utility in viral sequencing from complex clinical samples. Comparative genomics reveals the divergent strain clusters with the only completely sequenced HPIV4a subtype. However, it also exhibits various structural features present in one of the HPIV4b reference strains, opening questions regarding their lifecycle and evolutionary relationships among these viruses. Clinical data from patients infected with the strain, as well as viral prevalence estimates using real-time PCR, is also described.

DAFS: simultaneous aligning and folding of RNA sequences via dual decomposition

MOTIVATION

It is well known that the accuracy of RNA secondary structure prediction from a single sequence is limited, and thus a comparative approach that predicts a common secondary structure from aligned sequences is a better choice if homologous sequences with reliable alignments are available. However, correct secondary structure information is needed to produce reliable alignments of RNA sequences. To tackle this dilemma, we require a fast and accurate aligner that takes structural information into consideration to yield reliable structural alignments, which are suitable for common secondary structure prediction.

RESULTS

We develop DAFS, a novel algorithm that simultaneously aligns and folds RNA sequences based on maximizing expected accuracy of a predicted common secondary structure and its alignment. DAFS decomposes the pairwise structural alignment problem into two independent secondary structure prediction problems and one pairwise (non-structural) alignment problem by the dual decomposition technique, and maintains the consistency of a pairwise structural alignment by imposing penalties on inconsistent base pairs and alignment columns that are iteratively updated. Furthermore, we extend DAFS to consider pseudoknots in RNA structural alignments by integrating IPknot for predicting a pseudoknotted structure. The experiments on publicly available datasets showed that DAFS can produce reliable structural alignments from unaligned sequences in terms of accuracy of common secondary structure prediction.

Computational analysis of noncoding RNAs

Noncoding RNAs have emerged as important key players in the cell. Understanding their surprisingly diverse range of functions is challenging for experimental and computational biology. Here, we review computational methods to analyze noncoding RNAs. The topics covered include basic and advanced techniques to predict RNA structures, annotation of noncoding RNAs in genomic data, mining RNA-seq data for novel transcripts and prediction of transcript structures, computational aspects of microRNAs, and database resources.

A classification of bioinformatics algorithms from the viewpoint of maximizing expected accuracy (MEA)

Many estimation problems in bioinformatics are formulated as point estimation problems in a high-dimensional discrete space. In general, it is difficult to design reliable estimators for this type of problem, because the number of possible solutions is immense, which leads to an extremely low probability for every solution-even for the one with the highest probability. Therefore, maximum score and maximum likelihood estimators do not work well in this situation although they are widely employed in a number of applications. Maximizing expected accuracy (MEA) estimation, in which accuracy measures of the target problem and the entire distribution of solutions are considered, is a more successful approach. In this review, we provide an extensive discussion of algorithms and software based on MEA. We describe how a number of algorithms used in previous studies can be classified from the viewpoint of MEA. We believe that this review will be useful not only for users wishing to utilize software to solve the estimation problems appearing in this article, but also for developers wishing to design algorithms on the basis of MEA.

Understanding the transcriptome through RNA structure

RNA structure is crucial for gene regulation and function. In the past, transcriptomes have largely been parsed by primary sequences and expression levels, but it is now becoming feasible to annotate and compare transcriptomes based on RNA structure. In addition to computational prediction methods, the recent advent of experimental techniques to probe RNA structure by high-throughput sequencing has enabled genome-wide measurements of RNA structure and has provided the first picture of the structural organization of a eukaryotic transcriptome - the 'RNA structurome'. With additional advances in method refinement and interpretation, structural views of the transcriptome should help to identify and validate regulatory RNA motifs that are involved in diverse cellular processes and thereby increase understanding of RNA function.

From structure prediction to genomic screens for novel non-coding RNAs

Non-coding RNAs (ncRNAs) are receiving more and more attention not only as an abundant class of genes, but also as regulatory structural elements (some located in mRNAs). A key feature of RNA function is its structure. Computational methods were developed early for folding and prediction of RNA structure with the aim of assisting in functional analysis. With the discovery of more and more ncRNAs, it has become clear that a large fraction of these are highly structured. Interestingly, a large part of the structure is comprised of regular Watson-Crick and GU wobble base pairs. This and the increased amount of available genomes have made it possible to employ structure-based methods for genomic screens. The field has moved from folding prediction of single sequences to computational screens for ncRNAs in genomic sequence using the RNA structure as the main characteristic feature. Whereas early methods focused on energy-directed folding of single sequences, comparative analysis based on structure preserving changes of base pairs has been efficient in improving accuracy, and today this constitutes a key component in genomic screens. Here, we cover the basic principles of RNA folding and touch upon some of the concepts in current methods that have been applied in genomic screens for de novo RNA structures in searches for novel ncRNA genes and regulatory RNA structure on mRNAs. We discuss the strengths and weaknesses of the different strategies and how they can complement each other.

RNAG: a new Gibbs sampler for predicting RNA secondary structure for unaligned sequences

MOTIVATION

RNA secondary structure plays an important role in the function of many RNAs, and structural features are often key to their interaction with other cellular components. Thus, there has been considerable interest in the prediction of secondary structures for RNA families. In this article, we present a new global structural alignment algorithm, RNAG, to predict consensus secondary structures for unaligned sequences. It uses a blocked Gibbs sampling algorithm, which has a theoretical advantage in convergence time. This algorithm iteratively samples from the conditional probability distributions P(Structure | Alignment) and P(Alignment | Structure). Not surprisingly, there is considerable uncertainly in the high-dimensional space of this difficult problem, which has so far received limited attention in this field. We show how the samples drawn from this algorithm can be used to more fully characterize the posterior space and to assess the uncertainty of predictions.

RESULTS

Our analysis of three publically available datasets showed a substantial improvement in RNA structure prediction by RNAG over extant prediction methods. Additionally, our analysis of 17 RNA families showed that the RNAG sampled structures were generally compact around their ensemble centroids, and at least 11 families had at least two well-separated clusters of predicted structures. In general, the distance between a reference structure and our predicted structure was large relative to the variation among structures within an ensemble.

AVAILABILITY

The Perl implementation of the RNAG algorithm and the data necessary to reproduce the results described in Sections 3.1 and 3.2 are available at http://ccmbweb.ccv.brown.edu/rnag.html

CONTACT

charles_lawrence@brown.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Site-specific mobilization of vinyl chloride respiration islands by a mechanism common in Dehalococcoides

Background

Vinyl chloride is a widespread groundwater pollutant and Group 1 carcinogen. A previous comparative genomic analysis revealed that the vinyl chloride reductase operon, vcrABC, of Dehalococcoides sp. strain VS is embedded in a horizontally-acquired genomic island that integrated at the single-copy tmRNA gene, ssrA.

Results

We targeted conserved positions in available genomic islands to amplify and sequence four additional vcrABC -containing genomic islands from previously-unsequenced vinyl chloride respiring Dehalococcoides enrichments. We identified a total of 31 ssrA-specific genomic islands from Dehalococcoides genomic data, accounting for 47 reductive dehalogenase homologous genes and many other non-core genes. Sixteen of these genomic islands contain a syntenic module of integration-associated genes located adjacent to the predicted site of integration, and among these islands, eight contain vcrABC as genetic 'cargo'. These eight vcrABC -containing genomic islands are syntenic across their ~12 kbp length, but have two phylogenetically discordant segments that unambiguously differentiate the integration module from the vcrABC cargo. Using available Dehalococcoides phylogenomic data we estimate that these ssrA-specific genomic islands are at least as old as the Dehalococcoides group itself, which in turn is much older than human civilization.

Conclusions

The vcrABC -containing genomic islands are a recently-acquired subset of a diverse collection of ssrA-specific mobile elements that are a major contributor to strain-level diversity in Dehalococcoides, and may have been throughout its evolution. The high similarity between vcrABC sequences is quantitatively consistent with recent horizontal acquisition driven by ~100 years of industrial pollution with chlorinated ethenes.

The PETfold and PETcofold web servers for intra- and intermolecular structures of multiple RNA sequences

The function of non-coding RNA genes largely depends on their secondary structure and the interaction with other molecules. Thus, an accurate prediction of secondary structure and RNA–RNA interaction is essential for the understanding of biological roles and pathways associated with a specific RNA gene. We present web servers to analyze multiple RNA sequences for common RNA structure and for RNA interaction sites. The web servers are based on the recent PET (Probabilistic Evolutionary and Thermodynamic) models PETfold and PETcofold, but add user friendly features ranging from a graphical layer to interactive usage of the predictors. Additionally, the web servers provide direct access to annotated RNA alignments, such as the Rfam 10.0 database and multiple alignments of 16 vertebrate genomes with human. The web servers are freely available at: http://rth.dk/resources/petfold/

CentroidHomfold-LAST: accurate prediction of RNA secondary structure using automatically collected homologous sequences

Although secondary structure predictions of an individual RNA sequence have been widely used in a number of sequence analyses of RNAs, accuracy is still limited. Recently, we proposed a method (called 'CentroidHomfold'), which includes information about homologous sequences into the prediction of the secondary structure of the target sequence, and showed that it substantially improved the performance of secondary structure predictions. CentroidHomfold, however, forces users to prepare homologous sequences of the target sequence. We have developed a Web application (CentroidHomfold-LAST) that predicts the secondary structure of the target sequence using automatically collected homologous sequences. LAST, which is a fast and sensitive local aligner, and CentroidHomfold are employed in the Web application. Computational experiments with a commonly-used data set indicated that CentroidHomfold-LAST substantially outperformed conventional secondary structure predictions including CentroidFold and RNAfold.

Generalized centroid estimators in bioinformatics

In a number of estimation problems in bioinformatics, accuracy measures of the target problem are usually given, and it is important to design estimators that are suitable to those accuracy measures. However, there is often a discrepancy between an employed estimator and a given accuracy measure of the problem. In this study, we introduce a general class of efficient estimators for estimation problems on high-dimensional binary spaces, which represent many fundamental problems in bioinformatics. Theoretical analysis reveals that the proposed estimators generally fit with commonly-used accuracy measures (e.g. sensitivity, PPV, MCC and F-score) as well as it can be computed efficiently in many cases, and cover a wide range of problems in bioinformatics from the viewpoint of the principle of maximum expected accuracy (MEA). It is also shown that some important algorithms in bioinformatics can be interpreted in a unified manner. Not only the concept presented in this paper gives a useful framework to design MEA-based estimators but also it is highly extendable and sheds new light on many problems in bioinformatics.

Bioinformatics for RNomics

Rapid improvements in high-throughput experimental technologies make it nowadays possible to study the expression, as well as changes in expression, of whole transcriptomes under different environmental conditions in a detailed view. We describe current approaches to identify genome-wide functional RNA transcripts (experimentally as well as computationally), and focus on computational methods that may be utilized to disclose their function. While genome databases offer a wealth of information about known and putative functions for protein-coding genes, functional information for novel non-coding RNA genes is almost nonexistent. This is mainly explained by the lack of established software tools to efficiently reveal the function and evolutionary origin of non-coding RNA genes. Here, we describe in detail computational approaches one may follow to annotate and classify an RNA transcript.