Prediction and verification of microRNA targets by MovingTargets, a highly adaptable prediction method

Gra-CRC-miRTar: The pre-trained nucleotide-to-graph neural networks to identify potential miRNA targets in colorectal cancer

Colorectal cancer (CRC) is the third most diagnosed cancer and the second deadliest cancer worldwide representing a major public health problem. In recent years, increasing evidence has shown that microRNA (miRNA) can control the expression of targeted human messenger RNA (mRNA) by reducing their abundance or translation, acting as oncogenes or tumor suppressors in various cancers, including CRC. Due to the significant up-regulation of oncogenic miRNAs in CRC, elucidating the underlying mechanism and identifying dysregulated miRNA targets may provide a basis for improving current therapeutic interventions. In this paper, we proposed Gra-CRC-miRTar, a pre-trained nucleotide-to-graph neural network framework, for identifying potential miRNA targets in CRC. Different from previous studies, we constructed two pre-trained models to encode RNA sequences and transformed them into de Bruijn graphs. We employed different graph neural networks to learn the latent representations. The embeddings generated from de Bruijn graphs were then fed into a Multilayer Perceptron (MLP) to perform the prediction tasks. Our extensive experiments show that Gra-CRC-miRTar achieves better performance than other deep learning algorithms and existing predictors. In addition, our analyses also successfully revealed 172 out of 201 functional interactions through experimentally validated miRNA-mRNA pairs in CRC. Collectively, our effort provides an accurate and efficient framework to identify potential miRNA targets in CRC, which can also be used to reveal miRNA target interactions in other malignancies, facilitating the development of novel therapeutics. The Gra-CRC-miRTar web server can be found at: http://gra-crc-mirtar.com/.

Gra-CRC-miRTar: The pre-trained nucleotide-to-graph neural networks to identify potential miRNA targets in colorectal cancer

Colorectal cancer (CRC) is the third most diagnosed cancer and the second deadliest cancer worldwide representing a major public health problem. In recent years, increasing evidence has shown that microRNA (miRNA) can control the expression of targeted human messenger RNA (mRNA) by reducing their abundance or translation, acting as oncogenes or tumor suppressors in various cancers, including CRC. Due to the significant up-regulation of oncogenic miRNAs in CRC, elucidating the underlying mechanism and identifying dysregulated miRNA targets may provide a basis for improving current therapeutic interventions. In this paper, we proposed Gra-CRC-miRTar, a pre-trained nucleotide-to-graph neural network framework, for identifying potential miRNA targets in CRC. Different from previous studies, we constructed two pre-trained models to encode RNA sequences and transformed them into de Bruijn graphs. We employed different graph neural networks to learn the latent representations. The embeddings generated from de Bruijn graphs were then fed into a Multilayer Perceptron (MLP) to perform the prediction tasks. Our extensive experiments show that Gra-CRC-miRTar achieves better performance than other deep learning algorithms and existing predictors. In addition, our analyses also successfully revealed 172 out of 201 functional interactions through experimentally validated miRNA-mRNA pairs in CRC. Collectively, our effort provides an accurate and efficient framework to identify potential miRNA targets in CRC, which can also be used to reveal miRNA target interactions in other malignancies, facilitating the development of novel therapeutics.

GraphTar: applying word2vec and graph neural networks to miRNA target prediction

BACKGROUND

MicroRNAs (miRNAs) are short, non-coding RNA molecules that regulate gene expression by binding to specific mRNAs, inhibiting their translation. They play a critical role in regulating various biological processes and are implicated in many diseases, including cardiovascular, oncological, gastrointestinal diseases, and viral infections. Computational methods that can identify potential miRNA-mRNA interactions from raw data use one-dimensional miRNA-mRNA duplex representations and simple sequence encoding techniques, which may limit their performance.

RESULTS

We have developed GraphTar, a new target prediction method that uses a novel graph-based representation to reflect the spatial structure of the miRNA-mRNA duplex. Unlike existing approaches, we use the word2vec method to accurately encode RNA sequence information. In conjunction with the novel encoding method, we use a graph neural network classifier that can accurately predict miRNA-mRNA interactions based on graph representation learning. As part of a comparative study, we evaluate three different node embedding approaches within the GraphTar framework and compare them with other state-of-the-art target prediction methods. The results show that the proposed method achieves similar performance to the best methods in the field and outperforms them on one of the datasets.

CONCLUSIONS

In this study, a novel miRNA target prediction approach called GraphTar is introduced. Results show that GraphTar is as effective as existing methods and even outperforms them in some cases, opening new avenues for further research. However, the expansion of available datasets is critical for advancing the field towards real-world applications.

Prediction methods for microRNA targets in bilaterian animals: Toward a better understanding by biologists

MicroRNAs (miRNAs) are small noncoding RNAs that regulate gene expression at the posttranscriptional level. Because of their wide network of interactions, miRNAs have become the focus of many studies over the past decade, particularly in animal species. To streamline the number of potential wet lab experiments, the use of miRNA target prediction tools is currently the first step undertaken. However, the predictions made may vary considerably depending on the tool used, which is mostly due to the complex and still not fully understood mechanism of action of miRNAs. The discrepancies complicate the choice of the tool for miRNA target prediction. To provide a comprehensive view of this issue, we highlight in this review the main characteristics of miRNA-target interactions in bilaterian animals, describe the prediction models currently used, and provide some insights for the evaluation of predictor performance.

miRTMC: A miRNA Target Prediction Method Based on Matrix Completion Algorithm

microRNAs (miRNAs) are small non-coding RNAs which modulate the stability of gene targets and their rates of translation into proteins at transcriptional level and post-transcriptional level. miRNA dysfunctions can lead to human diseases because of dysregulation of their targets. Correct miRNA target prediction will lead to better understanding of the mechanisms of human diseases and provide hints on curing them. In recent years, computational miRNA target prediction methods have been proposed according to the interaction rules between miRNAs and targets. However, these methods suffer from high false positive rates due to the complicated relationship between miRNAs and their targets. The rapidly growing number of experimentally validated miRNA targets enables predicting miRNA targets with high precision via accurate data analysis. Taking advantage of these known miRNA targets, a novel recommendation system model (miRTMC) for miRNA target prediction is established using a new matrix completion algorithm. In miRTMC, a heterogeneous network is constructed by integrating the miRNA similarity network, the gene similarity network, and the miRNA-gene interaction network. Our assumption is that the latent factors determining whether a gene is the target of miRNA or not are highly correlated, i.e., the adjacency matrix of the heterogeneous network is low-rank, which is then completed by using a nuclear norm regularized linear least squares model under non-negative constraints. Alternating direction method of multipliers (ADMM) is adopted to numerically solve the matrix completion problem. Our results show that miRTMC outperforms the competing methods in terms of various evaluation metrics. Our software package is available at https://github.com/hjiangcsu/miRTMC.

Soma influences GSC progeny differentiation via the cell adhesion-mediated steroid-let-7-Wingless signaling cascade that regulates chromatin dynamics

It is known that signaling from the germline stem cell niche is required to maintain germline stem cell identity in Drosophila. However, it is not clear whether the germline stem-cell daughters differentiate by default (because they are physically distant from the niche) or whether additional signaling is necessary to initiate the differentiation program. Previously, we showed that ecdysteroid signaling cell non-autonomously regulates early germline differentiation via its soma-specific co-activator and co-repressor, Taiman and Abrupt. Now, we demonstrate that this regulation is modulated by the miRNA let-7, which acts in a positive feedback loop to confer ecdysone signaling robustness via targeting its repressor, the transcription factor Abrupt. This feedback loop adjusts ecdysteroid signaling in response to some stressful alterations in the external and internal conditions, which include temperature stress and aging, but not nutritional deprivation. Upon let-7 deficit, escort cells fail to properly differentiate: their shape, division, and cell adhesive characteristics are perturbed. These cells have confused cellular identity and form columnar-like rather than squamous epithelium and fail to send protrusions in between differentiating germline cysts, affecting soma-germline communication. Particularly, levels of the homophilic cell adhesion protein Cadherin, which recruits Wg signaling transducer β-catenin, are increased in mutant escort cells and, correspondingly, in the adjacent germline cells. Readjustment of heterotypic (soma-germline) cell adhesion modulates Wg signaling intensity in the germline, which in turn regulates histone modifications that promote expression of the genes necessary to trigger early germline differentiation. Thus, our data first show the intrinsic role for Wg signaling in the germline and support a model where the soma influences the tempo of germline differentiation in response to external conditions.

Comprehensive analysis of microRNA genomic loci identifies pervasive repetitive-element origins

MicroRNAs (miRs) are small non-coding RNAs that generally function as negative regulators of target messenger RNAs (mRNAs) at the posttranscriptional level. MiRs bind to the 3'UTR of target mRNAs through complementary base pairing, resulting in target mRNA cleavage or translation repression. To date, over 15,000 distinct miRs have been identified in organisms ranging from viruses to man and interest in miR research continues to intensify. Of note, the most enlightening aspect of miR function-the mRNAs they target-continues to be elusive. Descriptions of the molecular origins of independent miR molecules currently support the hypothesis that miR hairpin generation is based on the adjacent insertion of two related transposable elements (TEs) at one genomic locus. Thus transcription across such TE interfaces establishes many, if not the majority of functional miRs. The implications of these findings are substantial for understanding how TEs confer increased genomic fitness, describing miR transcriptional regulations and making accurate miR target predictions. In this work, we have performed a comprehensive analysis of the genomic events responsible for the formation of all currently annotated miR loci. We find that the connection between miRs and transposable elements is more significant than previously appreciated, and more broadly, supports an important role for repetitive elements in miR origin, expression and regulatory network formation. Further, we demonstrate the utility of these findings in miR target prediction. Our results greatly expand the existing repertoire of defined miR origins, detailing the formation of 2,392 of 15,176 currently recognized miR genomic loci and supporting a mobile genetic element model for the genomic establishment of functional miRs.

Continuing analysis of microRNA origins: Formation from transposable element insertions and noncoding RNA mutations

MicroRNAs (miRs) are small noncoding RNAs that typically act as regulators of gene expression by base pairing with the 3' UTR of messenger RNAs (mRNAs) and either repressing their translation or initiating degradation. As of this writing over 24,500 distinct miRs have been identified, but the functions of the vast majority of these remain undescribed. This paper represents a summary of our in depth analysis of the genomic origins of miR loci, detailing the formation of 1,213 of the 7,321 recently identified miRs and thereby bringing the total number of miR loci with defined molecular origin to 3,605. Interestingly, our analyses also identify evidence for a second, novel mechanism of miR locus generation through describing the formation of 273 miR loci from mutations to other forms of noncoding RNAs. Importantly, several independent investigations of the genomic origins of miR loci have now supported the hypothesis that miR hairpins are formed by the adjacent genomic insertion of two complementary transposable elements (TEs) into opposing strands. While our results agree that subsequent transcription over such TE interfaces leads to the formation of the majority of functional miR loci, we now also find evidence suggesting that a subset of miR loci were actually formed by an alternative mechanism-point mutations in other structurally complex, noncoding RNAs (e.g., tRNAs and snoRNAs).

A Bayesian decision fusion approach for microRNA target prediction

MicroRNAs (miRNAs) are 19-25 nucleotides non-coding RNAs known to have important post-transcriptional regulatory functions. The computational target prediction algorithm is vital to effective experimental testing. However, since different existing algorithms rely on different features and classifiers, there is a poor agreement among the results of different algorithms. To benefit from the advantages of different algorithms, we proposed an algorithm called BCmicrO that combines the prediction of different algorithms with Bayesian Network. BCmicrO was evaluated using the training data and the proteomic data. The results show that BCmicrO improves both the sensitivity and the specificity of each individual algorithm. All the related materials including genome-wide prediction of human targets and a web-based tool are available at http://compgenomics.utsa.edu/gene/gene_1.php.

Steroid-induced microRNA let-7 acts as a spatio-temporal code for neuronal cell fate in the developing Drosophila brain

Mammalian neuronal stem cells produce multiple neuron types in the course of an individual's development. Similarly, neuronal progenitors in the Drosophila brain generate different types of closely related neurons that are born at specific time points during development. We found that in the post-embryonic Drosophila brain, steroid hormones act as temporal cues that specify the cell fate of mushroom body (MB) neuroblast progeny. Chronological regulation of neurogenesis is subsequently mediated by the microRNA (miRNA) let-7, absence of which causes learning impairment due to morphological MB defects. The miRNA let-7 is required to regulate the timing of α'/β' to α/β neuronal identity transition by targeting the transcription factor Abrupt. At a cellular level, the ecdysone-let-7-Ab signalling pathway controls the expression levels of the cell adhesion molecule Fasciclin II in developing neurons that ultimately influences their differentiation. Our data propose a novel role for miRNAs as transducers between chronologically regulated developmental signalling and physical cell adhesion.

Molecular dynamics simulations of Ago silencing complexes reveal a large repertoire of admissible 'seed-less' targets

To better understand the recognition mechanism of RISC and the repertoire of guide-target interactions we introduced G:U wobbles and mismatches at various positions of the microRNA (miRNA) 'seed' region and performed all-atom molecular dynamics simulations of the resulting Ago-miRNA:mRNA ternary complexes. Our simulations reveal that many modifications, including combinations of multiple G:U wobbles and mismatches in the seed region, are admissible and result in only minor structural fluctuations that do not affect overall complex stability. These results are further supported by analyses of HITS-CLIP data. Lastly, introduction of disruptive mutations revealed a bending motion of the PAZ domain along the L1/L2 'hinge' and a subsequent opening of the nucleic-acid-binding channel. Our findings suggest that the spectrum of a miRNA's admissible targets is different from what is currently anticipated by the canonical seed-model. Moreover, they provide a likely explanation for the previously reported sequence-dependent regulation of unintended targeting by siRNAs.

Small temporal RNAs in animal development

The lin-4/miR-125 and let-7 microRNAs are at the heart of the heterochronic pathway, which controls temporal cell fate determination during Caenorhabditis elegans development. These small temporal RNAs are clustered along with a third microRNA, miR-100, in the genomes of most animals. Their conserved temporal and neural expression profile suggests a general role in cell fate determination during nervous system differentiation. By triggering consecutive differentiation programs, these microRNAs probably help to determine birth-order dependent temporal identity and thereby contribute to neural stem cell multipotency.

An overview of microRNA methods: expression profiling and target identification

MicroRNAs (miRNAs) are small, single-stranded RNA molecules encoded by genes that are transcribed from DNA but not translated into protein (noncoding RNA). The ability of miRNA to regulate the expression of, as yet, an unknown quantity of targets has recently become an area of huge interest to researchers studying many different areas in many species. Identifying miRNA targets provides functional insights and strategies for therapy. Furthermore, the recent advent of high-throughput methods for profiling miRNA expression and for the identification of miRNA targets has ushered in a new era in the research of gene regulation. miRNA profiling further adds a new dimension of information for the molecular profiling of disease. Summarized herein are the methods used to query the expression of miRNAs at both an individual and global level. We have also described modern computational approaches to identifying miRNA target transcripts.

Evaluation of online miRNA resources for biomedical applications

MicroRNAs (miRNAs) are endogenous single-stranded, 22-nt (nucleotide) RNAs which complement mRNA to initiate post-transcriptional regulation. This review presents updates and evaluations of the public domain resources available for miRNA identification and target prediction toward their utilization in the biomedical research approach. This study discusses the basic principles of miRNA computational studies based on the nature and mechanism of action of miRNAs. Furthermore, we have explored fifty-nine current online miRNA tools that can be categorized into three classes in this paper: (i) miRNA identification; (ii) miRNA target prediction; and (iii) specialized miRNA tools.

Survey of Computational Algorithms for MicroRNA Target Prediction

MicroRNAs (miRNAs) are 19 to 25 nucleotides non-coding RNAs known to possess important post-transcriptional regulatory functions. Identifying targeting genes that miRNAs regulate are important for understanding their specific biological functions. Usually, miRNAs down-regulate target genes through binding to the complementary sites in the 3' untranslated region (UTR) of the targets. In part, due to the large number of miRNAs and potential targets, an experimental based prediction design would be extremely laborious and economically unfavorable. However, since the bindings of the animal miRNAs are not a perfect one-to-one match with the complementary sites of their targets, it is difficult to predict targets of animal miRNAs by accessing their alignment to the 3' UTRs of potential targets. Consequently, sophisticated computational approaches for miRNA target prediction are being considered as essential methods in miRNA research.

We surveyed most of the current computational miRNA target prediction algorithms in this paper. Particularly, we provided a mathematical definition and formulated the problem of target prediction under the framework of statistical classification. Moreover, we summarized the features of miRNA-target pairs in target prediction approaches and discussed these approaches according to two categories, which are the rule-based and the data-driven approaches. The rule-based approach derives the classifier mainly on biological prior knowledge and important observations from biological experiments, whereas the data driven approach builds statistic models using the training data and makes predictions based on the models. Finally, we tested a few different algorithms on a set of experimentally validated true miRNA-target pairs [1] and a set of false miRNA-target pairs, derived from miRNA overexpression experiment [2]. Receiver Operating Characteristic (ROC) curves were drawn to show the performances of these algorithms.

Identification of microRNA target genes in vivo

MicroRNAs (miRNAs) are short non-coding RNAs transcribed from intergenic or intronic sequences as long precursors that are sequentially processed by the endonucleases Drosha and Dicer into short double-stranded sequences. It is clear that miRNAs play essential roles in gene expression, development, and cell fate specification in animals. However, one of the barriers of miRNA research is how to find the target genes. In this study, we have developed a rapid and effective method to isolate miRNA target genes in vivo. MicroRNA was synthesized in vitro and labeled by biotin. After transfected into cells, the miRNA/mRNA complexes were isolated by streptavidin-coated magnetic beads. hsa-miR155 was taken as model to validate this method, which is a very important modulator in tumor development. It is useful for validation of targets predicted in silico, and, potentially, for discovery of previously uncharacterized targets.

MicroRNAs in Drosophila development

Micro-ribonucleic acids (miRNAs) are small (21-24 nucleotide), endogenously expressed, noncoding RNAs that have emerged as important posttranscriptional regulators of gene expression. MiRNAs have been identified and cloned from diverse eukaryotic organisms where they have been shown to control important physiological and developmental processes such as apoptosis, cell division, and differentiation. A high level of conservation of some miRNAs across phyla further emphasizes their importance as posttranscriptional regulators. Research in a variety of model systems has been instrumental in dissecting the biological functions of miRNAs. In this chapter, we discuss the current literature on the role of miRNAs as developmental regulators in Drosophila.

Viral miRNAs

Since 2004, more than 200 microRNAs (miRNAs) have been discovered in double-stranded DNA viruses, mainly herpesviruses and polyomaviruses (Nucleic Acids Res 32:D109-D111, 2004). miRNAs are short 22  ±  3 nt RNA molecules that posttranscriptionally regulate gene expression by binding to 3'-untranslated regions (3'UTR) of target mRNAs, thereby inducing translational silencing and/or transcript degradation (Nature 431:350-355, 2004; Cell 116:281-297, 2004). Since miRNAs require only limited complementarity for binding, miRNA targets are difficult to determine (Mol Cell 27:91-105, 2007). To date, targets have only been experimentally verified for relatively few viral miRNAs, which either target viral or host cellular gene expression: For example, SV40 and related polyomaviruses encode miRNAs which target viral large T antigen expression (Nature 435:682-686, 2005; J Virol 79:13094-13104, 2005; Virology 383:183-187, 2009; J Virol 82:9823-9828, 2008) and miRNAs of α-, β-, and γ-herpesviruses have been implicated in regulating the transition from latent to lytic gene expression, a key step in the herpesvirus life cycle. Viral miRNAs have also been shown to target various host cellular genes. Although this field is just beginning to unravel the multiple roles of viral miRNA in biology and pathogenesis, the current data strongly suggest that virally encoded miRNAs are able to regulate fundamental biological processes such as immune recognition, promotion of cell survival, angiogenesis, proliferation, and cell differentiation. This chapter aims to summarize our current knowledge of viral miRNAs, their targets and function, and the challenges lying ahead to decipher their role in viral biology, pathogenesis, and for γ-herepsvirus-encoded miRNAs, potentially tumorigenesis.

A Bayesian approach for identifying miRNA targets by combining sequence prediction and gene expression profiling

Background

MicroRNAs (miRNAs) are single-stranded non-coding RNAs shown to plays important regulatory roles in a wide range of biological processes and diseases. The functions and regulatory mechanisms of most of miRNAs are still poorly understood in part because of the difficulty in identifying the miRNA regulatory targets. To this end, computational methods have evolved as important tools for genome-wide target screening. Although considerable work in the past few years has produced many target prediction algorithms, most of them are solely based on sequence, and the accuracy is still poor. In contrast, gene expression profiling from miRNA transfection experiments can provide additional information about miRNA targets. However, most of existing research assumes down-regulated mRNAs as targets. Given the fact that the primary function of miRNA is protein inhibition, this assumption is neither sufficient nor necessary.

Results

A novel Bayesian approach is proposed in this paper that integrates sequence level prediction with expression profiling of miRNA transfection. This approach does not restrict the target to be down-expressed and thus improve the performance of existing target prediction algorithm. The proposed algorithm was tested on simulated data, proteomics data, and IP pull-down data and shown to achieve better performance than existing approaches for target prediction. All the related materials including source code are available at http://compgenomics.utsa.edu/expmicro.html.

Conclusions

The proposed Bayesian algorithm integrates properly the sequence paring data and mRNA expression profiles for miRNA target prediction. This algorithm is shown to have better prediction performance than existing algorithms.

Hormonal regulation of Drosophila microRNA let-7 and miR-125 that target innate immunity

The steroid 20-hydroxy-ecdysone (20-HE) and the sesquiterpenoid Juvenile Hormone (JH) coordinate insect life stage transitions. 20-HE exerts these effects by the sequential induction of response genes. In the nematode Caenorhabditis elegans hormones also play a role in such transitions, but notably, microRNA such as let-7 and lin-4 have likewise been found to help order developmental steps. Little is known about the corresponding function of homologous microRNA in Drosophila melanogaster, and the way microRNA might be regulated by 20-HE in the fly is ambiguous. Here we used Drosophila S2 cells to analyze the effects of 20-HE on D. melanogaster microRNA let-7 and miR-125, the homolog of lin-4. The induction by 20-HE of let-7 and miR-125 in S2 cells is inhibited by RNAi knockdown of the ecdysone receptor and, as previously shown, by knockdown of its cofactor broad-complex C. To help resolve the currently ambiguous role of 20-HE in the control of microRNA, we show that nanomolar concentrations of 20-HE primes cells to subsequently express microRNA when exposed to micromolar levels of 20-HE. We then explore the role microRNA plays in the established relationship between 20-HE and the induction of innate immunity. We show that the 3'UTR of the antimicrobial peptide diptericin has a let-7 binding site and that let-7 represses translation from this site. We conclude that 20-HE facilitates the initial expression of innate immunity while it simultaneously induces negative regulation via microRNA control of antimicrobial peptide translation.

Improving performance of mammalian microRNA target prediction

Background

MicroRNAs (miRNAs) are single-stranded non-coding RNAs known to regulate a wide range of cellular processes by silencing the gene expression at the protein and/or mRNA levels. Computational prediction of miRNA targets is essential for elucidating the detailed functions of miRNA. However, the prediction specificity and sensitivity of the existing algorithms are still poor to generate meaningful, workable hypotheses for subsequent experimental testing. Constructing a richer and more reliable training data set and developing an algorithm that properly exploits this data set would be the key to improve the performance current prediction algorithms.

Results

A comprehensive training data set is constructed for mammalian miRNAs with its positive targets obtained from the most up-to-date miRNA target depository called miRecords and its negative targets derived from 20 microarray data. A new algorithm SVMicrO is developed, which assumes a 2-stage structure including a site support vector machine (SVM) followed by a UTR-SVM. SVMicrO makes prediction based on 21 optimal site features and 18 optimal UTR features, selected by training from a comprehensive collection of 113 site and 30 UTR features. Comprehensive evaluation of SVMicrO performance has been carried out on the training data, proteomics data, and immunoprecipitation (IP) pull-down data. Comparisons with some popular algorithms demonstrate consistent improvements in prediction specificity, sensitivity and precision in all tested cases. All the related materials including source code and genome-wide prediction of human targets are available at http://compgenomics.utsa.edu/svmicro.html.

Conclusions

A 2-stage SVM based new miRNA target prediction algorithm called SVMicrO is developed. SVMicrO is shown to be able to achieve robust performance. It holds the promise to achieve continuing improvement whenever better training data that contain additional verified or high confidence positive targets and properly selected negative targets are available.

MicroRNAs--targeting and target prediction

MicroRNAs (miRNAs) are a class of small noncoding RNAs that can regulate many genes by base pairing to sites in mRNAs. The functionality of miRNAs overlaps that of short interfering RNAs (siRNAs), and many features of miRNA targeting have been revealed experimentally by studying miRNA-mimicking siRNAs. This review outlines the features associated with animal miRNA targeting and describes currently available prediction tools.

MTar: a computational microRNA target prediction architecture for human transcriptome

BACKGROUND

MicroRNAs (miRNAs) play an essential task in gene regulatory networks by inhibiting the expression of target mRNAs. As their mRNA targets are genes involved in important cell functions, there is a growing interest in identifying the relationship between miRNAs and their target mRNAs. So, there is now a imperative need to develop a computational method by which we can identify the target mRNAs of existing miRNAs. Here, we proposed an efficient machine learning model to unravel the relationship between miRNAs and their target mRNAs.

RESULTS

We present a novel computational architecture MTar for miRNA target prediction which reports 94.5% sensitivity and 90.5% specificity. We identified 16 positional, thermodynamic and structural parameters from the wet lab proven miRNA:mRNA pairs and MTar makes use of these parameters for miRNA target identification. It incorporates an Artificial Neural Network (ANN) verifier which is trained by wet lab proven microRNA targets. A number of hitherto unknown targets of many miRNA families were located using MTar. The method identifies all three potential miRNA targets (5' seed-only, 5' dominant, and 3' canonical) whereas the existing solutions focus on 5' complementarities alone.

CONCLUSION

MTar, an ANN based architecture for identifying functional regulatory miRNA-mRNA interaction using predicted miRNA targets. The area of target prediction has received a new momentum with the function of a thermodynamic model incorporating target accessibility. This model incorporates sixteen structural, thermodynamic and positional features of residues in miRNA: mRNA pairs were employed to select target candidates. So our novel machine learning architecture, MTar is found to be more comprehensive than the existing methods in predicting miRNA targets, especially human transcritome.

Current tools for the identification of miRNA genes and their targets

The discovery of microRNAs (miRNAs), almost 10 years ago, changed dramatically our perspective on eukaryotic gene expression regulation. However, the broad and important functions of these regulators are only now becoming apparent. The expansion of our catalogue of miRNA genes and the identification of the genes they regulate owe much to the development of sophisticated computational tools that have helped either to focus or interpret experimental assays. In this article, we review the methods for miRNA gene finding and target identification that have been proposed in the last few years. We identify some problems that current approaches have not yet been able to overcome and we offer some perspectives on the next generation of computational methods.

miRNA-dependent translational repression in the Drosophila ovary

Background

The Drosophila ovary is a tissue rich in post-transcriptional regulation of gene expression. Many of the regulatory factors are proteins identified via genetic screens. The more recent discovery of microRNAs, which in other animals and tissues appear to regulate translation of a large fraction of all mRNAs, raised the possibility that they too might act during oogenesis. However, there has been no direct demonstration of microRNA-dependent translational repression in the ovary.

Methodology/Principal Findings

Here, quantitative analyses of transcript and protein levels of transgenes with or without synthetic miR-312 binding sites show that the binding sites do confer translational repression. This effect is dependent on the ability of the cells to produce microRNAs. By comparison with microRNA-dependent translational repression in other cell types, the regulated mRNAs and the protein factors that mediate repression were expected to be enriched in sponge bodies, subcellular structures with extensive similarities to the P bodies found in other cells. However, no such enrichment was observed.

Conclusions/Significance

Our results reveal the variety of post-transcriptional regulatory mechanisms that operate in the Drosophila ovary, and have implications for the mechanisms of miRNA-dependent translational control used in the ovary.

miRNAs: from biogenesis to networks

In eukaryotes, besides alternative splicing and promoter regulation of "classical" genes, there is also another level of genetic regulation based on non-coding RNAs (ncRNAs). The most famous group of ncRNAs is microRNAs, probably the biggest number of genome regulators. Here, we summarize the knowledge that has been accumulated about the microRNA field, focusing our attention on brief history, biogenesis, regulated mechanism, computational methods of miRNA finding and miRNA target sites, miRNAs and diseases, and miRNAs and network analysis.

Developmental timing: let-7 function conserved through evolution

Expression of the heterochronic microRNA let-7 is tightly correlated with the onset of adult development in many animals, suggesting that it functions as an evolutionarily conserved developmental timer. This hypothesis has now been confirmed by studies in Drosophila.

Warts is required for PI3K-regulated growth arrest, autophagy, and autophagic cell death in Drosophila

BACKGROUND

Cell growth arrest and autophagy are required for autophagic cell death in Drosophila. Maintenance of growth by expression of either activated Ras, Dp110, or Akt is sufficient to inhibit autophagy and cell death in Drosophila salivary glands, but the mechanism that controls growth arrest is unknown. Although the Warts (Wts) tumor suppressor is a critical regulator of tissue growth in animals, it is not clear how this signaling pathway controls cell growth.

RESULTS

Here, we show that genes in the Wts pathway are required for salivary gland degradation and that wts mutants have defects in cell growth arrest, caspase activity, and autophagy. Expression of Atg1, a regulator of autophagy, in salivary glands is sufficient to rescue wts mutant salivary gland destruction. Surprisingly, expression of Yorkie (Yki) and Scalloped (Sd) in salivary glands fails to phenocopy wts mutants. By contrast, misexpression of the Yki target bantam was able to inhibit salivary gland cell death, even though mutations in bantam fail to suppress the wts mutant salivary gland-persistence phenotype. Significantly, wts mutant salivary glands possess altered phosphoinositide signaling, and decreased function of the class I PI3K-pathway genes chico and TOR suppressed wts defects in cell death.

CONCLUSIONS

Although we have previously shown that salivary gland degradation requires genes in the Wts pathway, this study provides the first evidence that Wts influences autophagy. Our data indicate that the Wts-pathway components Yki, Sd, and bantam fail to function in salivary glands and that Wts regulates salivary gland cell death in a PI3K-dependent manner.

Kallikreins as microRNA targets: an in silico and experimental-based analysis

microRNAs (miRNAs) are non-coding RNAs that target specific mRNAs. They have been shown to control many biological processes including cancer pathogenesis. Kallikreins (KLKs) are a family of serine proteases that are attracting interest as cancer biomarkers. The mechanism of regulation of kallikrein expression is largely unknown. We investigated the potential roles of miRNAs in regulating KLK expression. Using a bioinformatics approach, we identified 96 strong KLK/miRNA interactions. KLK10 is the most frequently targeted kallikrein, followed by KLK5 and KLK13. KLK1, KLK3, KLK8 and KLK12 do not have strongly predicted miRNA/KLK interactions. Ten miRNAs are predicted to target more than one KLK. KLK2, KLK4, KLK5 and KLK10 have multiple miRNA-targeting sites on their transcript. Chromosomes 19 and 14 harbor significantly more KLK-targeting miRNAs. Many KLK-targeting miRNAs have been shown to be dysregulated in malignancy. We experimentally verified our bioinformatics data for the let-7f miRNA in a cell line model. let-7f transfection led to a significant decrease in secreted KLK6 and KLK10 protein levels. Co-transfection of let-7f and anti-let-7f inhibitor was able to partially rescue these protein levels. We conclude that miRNAs play a role in the regulation of KLK expression. Further studies are needed to investigate whether this regulation is altered in cancer.

Trans-natural antisense transcripts including noncoding RNAs in 10 species: implications for expression regulation

Natural antisense transcripts are at least partially complementary to their sense transcripts. Cis-Sense/Antisense pairs (cis-SAs) have been extensively characterized and known to play diverse regulatory roles, whereas trans-Sense/Antisense pairs (trans-SAs) in animals are poorly studied. We identified long trans-SAs in human and nine other animals, using ESTs to increase coverage significantly over previous studies. The percentage of transcriptional units (TUs) involved in trans-SAs among all TUs was as high as 4.13%. Particularly 2896 human TUs (or 2.89% of all human TUs) were involved in 3327 trans-SAs. Sequence complementarities over multiple segments with predicted RNA hybridization indicated that some trans-SAs might have sophisticated RNA-RNA pairing patterns. One-fourth of human trans-SAs involved noncoding TUs, suggesting that many noncoding RNAs may function by a trans-acting antisense mechanism. TUs in trans-SAs were statistically significantly enriched in nucleic acid binding, ion/protein binding and transport and signal transduction functions and pathways; a significant number of human trans-SAs showed concordant or reciprocal expression pattern; a significant number of human trans-SAs were conserved in mouse. This evidence suggests important regulatory functions of trans-SAs. In 30 cases, trans-SAs were related to cis-SAs through paralogues, suggesting a possible mechanism for the origin of trans-SAs. All trans-SAs are available at http://trans.cbi.pku.edu.cn/.

Temporal regulation of metamorphic processes in Drosophila by the let-7 and miR-125 heterochronic microRNAs

BACKGROUND

The let-7 and lin-4 microRNAs belong to a class of temporally expressed, noncoding regulatory RNAs that function as heterochronic switch genes in the nematode C. elegans. Heterochronic genes control the relative timing of events during development and are considered a major force in the rapid evolution of new morphologies. let-7 is highly conserved and in Drosophila is temporally coregulated with the lin-4 homolog, miR-125. Little is known, however, about their requirement outside the nematode or whether they universally control the timing of developmental processes.

RESULTS

We report the generation of a Drosophila mutant that lacks let-7 and miR-125 activities and that leads to a pleiotropic phenotype arising during metamorphosis. We focus on two defects and demonstrate that loss of let-7 and miR-125 results in temporal delays in two distinct metamorphic processes: the terminal cell-cycle exit in the wing and maturation of neuromuscular junctions (NMJs) at adult abdominal muscles. We identify the abrupt (ab) gene, encoding a nuclear protein, as a bona fide let-7 target and provide evidence that let-7 governs the maturation rate of abdominal NMJs during metamorphosis by regulating ab expression.

CONCLUSIONS

Drosophila Iet-7 and miR-125 mutants exhibit temporal misregulation of specific metamorphic processes. As in C. elegans, Drosophila let-7 is both necessary and sufficient for the appropriate timing of a specific cell-cycle exit, indicating that its function as a heterochronic microRNA is conserved. The ab gene is a target of let-7, and its repression in muscle is essential for the timing of NMJ maturation during metamorphosis. Our results suggest that let-7 and miR-125 serve as conserved regulators of events necessary for the transition from juvenile to adult life stages.

Regulation of the Drosophila lin-41 homologue dappled by let-7 reveals conservation of a regulatory mechanism within the LIN-41 subclade

Drosophila Dappled (DPLD) is a member of the RBCC/TRIM superfamily, a protein family involved in numerous diverse processes such as developmental timing and asymmetric cell divisions. DPLD belongs to the LIN-41 subclade, several members of which are micro RNA (miRNA) regulated. We re-examined the LIN-41 subclade members and their relation to other RBCC/TRIMs and dpld paralogs, and identified a new Drosophila muscle specific RBCC/TRIM: Another B-Box Affiliate, ABBA. In silico predictions of candidate miRNA regulators of dpld identified let-7 as the strongest candidate. Overexpression of dpld led to abnormal eye development, indicating that strict regulation of dpld mRNA levels is crucial for normal eye development. This phenotype was sensitive to let-7 dosage, suggesting let-7 regulation of dpld in the eye disc. A cell-based assay verified let-7 miRNA down-regulation of dpld expression by means of its 3'-untranslated region. Thus, dpld seems also to be miRNA regulated, suggesting that miRNAs represent an ancient mechanism of LIN-41 regulation.

microRNAs: a new frontier in kallikrein research

microRNAs (miRNAs) are a recently discovered class of small non-coding RNAs that regulate gene expression. Rapidly accumulating evidence has revealed that miRNAs are associated with cancer. The human tissue kalli-krein gene family is the largest contiguous family of proteases in the human genome, containing 15 genes. Many kallikreins have been reported as potential tumor markers. In this review, recent bioinformatics and experimental evidence is presented indicating that kallikreins are potential miRNA targets. The available experimental approaches to investigate these interactions and the potential diagnostic and therapeutic applications are also discussed. miRNAs represent a possible regulatory mechanism for controlling kallikrein expression at the post-transcriptional level. Many miRNAs were predicted to target kallikreins and a single miRNA can target more than one kallikrein. Recent evidence suggests that miRNAs can also exert ‘quantitative’ control of kallikreins by utilizing multiple targeting sites in the kallikrein mRNA. More research is needed to experimentally verify the in silico predictions and to investigate the possible role in tumor initiation and/or progression.

A scoring matrix approach to detecting miRNA target sites

BACKGROUND

Experimental identification of microRNA (miRNA) targets is a difficult and time consuming process. As a consequence several computational prediction methods have been devised in order to predict targets for follow up experimental validation. Current computational target prediction methods use only the miRNA sequence as input. With an increasing number of experimentally validated targets becoming available, utilising this additional information in the search for further targets may help to improve the specificity of computational methods for target site prediction.

RESULTS

We introduce a generic target prediction method, the Stacking Binding Matrix (SBM) that uses both information about the miRNA as well as experimentally validated target sequences in the search for candidate target sequences. We demonstrate the utility of our method by applying it to both animal and plant data sets and compare it with miRanda, a commonly used target prediction method.

CONCLUSION

We show that SBM can be applied to target prediction in both plants and animals and performs well in terms of sensitivity and specificity. Open source code implementing the SBM method, together with documentation and examples are freely available for download from the address in the Availability and Requirements section.

An overview of the identification, detection, and functional analysis of Drosophila microRNAs

MicroRNAs (miRNAs), small noncoding RNAs that post-transcriptionally regulate gene expression, are one of the most abundant classes of gene regulators. Yet, little is known about the roles that specific miRNAs play in the development of multicellular organisms. Drosophila provides an excellent model system to explore the in vivo activities of particular miRNAs within the context of well-defined gene-expression programs that control the development of a complex organism. This chapter reviews the various approaches currently used to identify Drosophila miRNAs, detect their expression, determine their messenger RNA targets, and study their function.

MicroRNA detection and target prediction: integration of computational and experimental approaches

In recent years, microRNAs (miRNAs), a class of 19-25 nucleotides noncoding RNAs, have been shown to play a major role in gene regulation across a broad range of metazoans and are important for a diverse biological functions. These miRNAs are involved in the regulation of protein expression primarily by binding to one or more target sites on an mRNA transcript and causing cleavage or repression of translation. Computer-based approaches for miRNA gene identification and miRNA target prediction are being considered as indispensable in miRNA research. Similarly, effective experimental techniques validating in silico predictions are crucial to the testing and finetuning of computational algorithms. Iterative interactions between in silico and experimental methods are now playing a central role in the biology of miRNAs. In this review, we summarize the various computational methods for identification of miRNAs and their targets as well as the technologies that have been developed to validate the predictions.

Bio-informatic trends for the determination of miRNA-target interactions in mammals

MicroRNAs (miRNAs) are small non-coding RNA molecules that regulate mRNAs through a sequence-specific mechanism. By virtue of their structure and mechanism of action, computational methods have been devised to investigate the encoding of miRNA genes and the targets of miRNA action. A variety of assumptions have predicated the implementation of these various computational solutions. Evolutionary sequence conservation, secondary structure, and folding energetics are some of the assumptions that have been used. The success of these different computational solutions has been evaluated for both elucidation of new miRNAs and deducing targets of miRNA action. While the focus is on search techniques for new miRNAs, we have compared the programs miRseeker, miRScan, PalGrade, ProMiR, and miRAlign as examples of implementation of these techniques. For these programs, a benchmark comparison between theoretical estimation and actual identification is possible. We have also compared the target prediction programs TargetScanS, PicTar, DIANA-microT, miRanda, and RNAhybrid. However, it is difficult to rigorously assess the benchmark performance of these programs due to the difficulty in confirming their theoretical predictions.

Regulatory small RNAs in plants

The discovery of microRNAs in the last decade altered the paradigm that protein coding genes are the only significant components for the regulation of gene networks. Within a short period of time small RNA systems within regulatory networks of eukaryotic cells have been uncovered that will ultimately change the way we infer gene regulation networks from transcriptional profiling data. Small RNAs are involved in the regulation of global activities of genic regions via chromatin states, as inhibitors of 'selfish' sequences (transposons, retroviruses), in establishment or maintenance of tissue/organ identity, and as modulators of the activity of transcription factor as well as 'house keeping' genes. With this chapter we provide an overview of the central aspects of small RNA function in plants and the features that distinguish the different small RNAs. We furthermore highlight the use of computational prediction methods for identification of plant miRNAs/precursors and their targets and provide examples for the experimental validation of small RNA candidates that could represent trans-regulators of downstream genes. Lastly, the emerging concepts of small RNAs as modulators of gene expression constituting systems networks within different cells in a multicellular organism are discussed.

Algorithms for mapping of mRNA targets for microRNA

MicroRNAs (miRNAs) are involved in human health and disease as endogenous suppressors of the translation of coding genes. At this early point of time in miRNA biology, it is important to identify specific cognate mRNA targets for miRNA. Investigation of the significance of miRNAs in disease processes relies on algorithms that hypothetically link specific miRNAs to their putative target genes. The development of such algorithms represents a hot area of research in biomedical informatics. Lack of biological data linking specific miRNAs to their respective mRNA targets represents the most serious limitation at this time. This article presents a concise review addressing the most popular concepts underlying state-of-the-art algorithms and principles aimed at target mapping for specific miRNAs. Strategies for improvement of the current bioinformatics tools and effective approaches for biological validation are discussed.

RNAs everywhere: genome-wide annotation of structured RNAs

Starting with the discovery of microRNAs and the advent of genome-wide transcriptomics, non-protein-coding transcripts have moved from a fringe topic to a central field research in molecular biology. In this contribution we review the state of the art of "computational RNomics", i.e., the bioinformatics approaches to genome-wide RNA annotation. Instead of rehashing results from recently published surveys in detail, we focus here on the open problem in the field, namely (functional) annotation of the plethora of putative RNAs. A series of exploratory studies are used to provide non-trivial examples for the discussion of some of the difficulties.

Inference of miRNA targets using evolutionary conservation and pathway analysis

BACKGROUND

MicroRNAs have emerged as important regulatory genes in a variety of cellular processes and, in recent years, hundreds of such genes have been discovered in animals. In contrast, functional annotations are available only for a very small fraction of these miRNAs, and even in these cases only partially.

RESULTS

We developed a general Bayesian method for the inference of miRNA target sites, in which, for each miRNA, we explicitly model the evolution of orthologous target sites in a set of related species. Using this method we predict target sites for all known miRNAs in flies, worms, fish, and mammals. By comparing our predictions in fly with a reference set of experimentally tested miRNA-mRNA interactions we show that our general method performs at least as well as the most accurate methods available to date, including ones specifically tailored for target prediction in fly. An important novel feature of our model is that it explicitly infers the phylogenetic distribution of functional target sites, independently for each miRNA. This allows us to infer species-specific and clade-specific miRNA targeting. We also show that, in long human 3' UTRs, miRNA target sites occur preferentially near the start and near the end of the 3' UTR. To characterize miRNA function beyond the predicted lists of targets we further present a method to infer significant associations between the sets of targets predicted for individual miRNAs and specific biochemical pathways, in particular those of the KEGG pathway database. We show that this approach retrieves several known functional miRNA-mRNA associations, and predicts novel functions for known miRNAs in cell growth and in development.

CONCLUSION

We have presented a Bayesian target prediction algorithm without any tunable parameters, that can be applied to sequences from any clade of species. The algorithm automatically infers the phylogenetic distribution of functional sites for each miRNA, and assigns a posterior probability to each putative target site. The results presented here indicate that our general method achieves very good performance in predicting miRNA target sites, providing at the same time insights into the evolution of target sites for individual miRNAs. Moreover, by combining our predictions with pathway analysis, we propose functions of specific miRNAs in nervous system development, inter-cellular communication and cell growth. The complete target site predictions as well as the miRNA/pathway associations are accessible on the ElMMo web server.

Computational Methods for MicroRNA Target Prediction

The discovery of microRNAs (miRNAs) has introduced a new paradigm into gene regulatory systems. Large numbers of miRNAs have been identified in a wide range of species, and most of them are known to downregulate translation of messenger RNAs (mRNAs) via imperfect binding of the miRNA to a specific site or sites in the 3' untranslated region (UTR) of the mRNA. Identification of genes targeted by miRNAs is widely believed to be an important step toward understanding the role of miRNAs in gene regulatory networks. As part of the effort to understand interactions between miRNAs and their targets, computational algorithms have been developed based on observed rules for features such as the degree of hybridization between the two RNA molecules. These in silico approaches provide important tools for miRNA target detection, and together with experimental validation, help to reveal regulated targets of miRNAs. Here, we summarize the knowledge that has been accumulated about the principles of target recognition by miRNAs and the currently available computational methodologies for prediction of miRNA target genes.

Computational identification of microRNAs and their targets

MicroRNAs (miRNAs) are one class of newly identified riboregulators of gene expression in many eukaryotic organisms. They play important roles in multiple biological and metabolic processes, including developmental timing, signal transduction, cell maintenance and differentiation, diseases and cancers. miRNAs regulate gene expression at the posttranscriptional level by directly cleaving targeted mRNAs or repressing translation. Although the founding members of miRNAs were discovered by genetic screening approaches, experimental approaches were limited by their low efficiency, time consuming, and high cost. As an alternative, computational approaches were developed. Computational approaches for identifying miRNAs are based on the following major characteristics of miRNAs: hairpin-shaped secondary structures, high conservation for some miRNAs, and high minimal folding free energy index (MFEI). Computational approaches also play an important role in identifying miRNA targets. A majority of known miRNAs and their targets were identified by computational approaches. Several web-based or non-web-based computer software programs are publicly available for predicting miRNAs and their targets.

Computational identification of microRNAs and their targets

One of the most important advances in biology in recent years may be the discovery of RNAs that can regulate gene expression. As one kind of such functional noncoding RNAs, microRNAs (miRNAs) form a class of endogenous 19-23-nucleotide RNAs that can have important regulatory roles in animals and plants by targeting transcripts for cleavage or translational repression. Since the discovery of the very first miRNAs, computational methods have been an invaluable tool that can complement experimental approaches to understand the biology of miRNAs. Most computational approaches associated with miRNA research can be classified into two broad categories, namely miRNA gene identification and miRNA target prediction. In this review, we summarize the principles of in silico prediction of miRNAs and their targets, and provide a comprehensive survey of specific methods that have been proposed in the field.

microRNA target predictions in animals

In recent years, microRNAs (miRNAs) have emerged as a major class of regulatory genes, present in most metazoans and important for a diverse range of biological functions. Because experimental identification of miRNA targets is difficult, there has been an explosion of computational target predictions. Although the initial round of predictions resulted in very diverse results, subsequent computational and experimental analyses suggested that at least a certain class of conserved miRNA targets can be confidently predicted and that this class of targets is large, covering, for example, at least 30% of all human genes when considering about 60 conserved vertebrate miRNA gene families. Most recent approaches have also shown that there are correlations between domains of miRNA expression and mRNA levels of their targets. Our understanding of miRNA function is still extremely limited, but it may be that by integrating mRNA and miRNA sequence and expression data with other comparative genomic data, we will be able to gain global and yet specific insights into the function and evolution of a broad layer of post-transcriptional control.

A Novel Method to Detect Functional MicroRNA Targets

MicroRNA (miRNA) molecules are non-coding RNAs, 19 to 24 nt in length that have been identified recently as important regulators of gene expression. Several computational methods have been developed to describe the target recognition mechanism by miRNA. We propose here a novel method to detect miRNA-mRNA complexes in eukaryotic cells. As a first step, we synthesize cDNA on an mRNA template using miRNAs as the endogenous cytoplasmic primer. This step extends miRNA and overcomes the problem of low complementary binding of miRNAs to their targets. Purified hybrid 3'-cDNA-miRNA-5' molecules are used in a second round of reverse transcription to anneal to target mRNA in a highly gene-specific manner. The 5'-end analysis of these cDNA molecules demonstrated that primers for cDNAs were "signatures" of miRNA molecules, and over-expression of their full-length mature miRNAs resulted in functional inhibition of target protein expression.

Virus-encoded microRNAs: a new chapter in virus–host cell interactions

MicroRNAs (miRNAs) have emerged as a paradigm shift in how gene regulation is viewed. Only 21–24 bases long, these noncoding RNAs regulate gene expression post-transcriptionally by binding to target genes and silencing their expression. Originally discovered in Caenorhabditis elegans in 1993, miRNAs have been found in a wide variety of organisms including nematodes, mammals and plants. In 2004, Epstein–Barr virus, was the first virus shown to encode and express miRNAs. Recently, miRNAs have also been cloned from Kaposi’s sarcoma-associated herpesvirus, mouse γ-herpesvirus-68 and human cytomegalovirus. In addition, SV40 and SA12 both express two miRNAs from a single hairpin, and the SV40 miRNA have been shown to negatively regulate large T-antigen expression. This review will summarize recent findings on the discovery, prediction, expression and potential functions of virally encoded miRNAs. Finally, some hypotheses on how these novel regulators may contribute to viral biology will also be presented.