The characterization of microRNA-mediated gene regulation as impacted by both target site location and seed match type

Network based multifactorial modelling of miRNA-target interactions

Competing endogenous RNA (ceRNA) regulations and crosstalk between various types of non-coding RNA in humans is an important and under-explored subject. Several studies have pointed out that an alteration in miRNA:target interaction can result in unexpected changes due to indirect and complex interactions. In this article, we defined a new network-based model that incorporates miRNA:ceRNA interactions with expression values. Our approach calculates network-wide effects of perturbations in the expression level of one or more nodes in the presence or absence of miRNA interaction factors such as seed type, binding energy. We carried out the analysis of large-scale miRNA:target networks from breast cancer patients. Highly perturbing genes identified by our approach coincide with breast cancer-associated genes and miRNAs. Our network-based approach takes the sponge effect into account and helps to unveil the crosstalk between nodes in miRNA:target network. The model has potential to reveal unforeseen regulations that are only evident in the network context. Our tool is scalable and can be plugged in with emerging miRNA effectors such as circRNAs, lncRNAs, and available as R package ceRNAnetsim: https://www.bioconductor.org/packages/release/bioc/html/ceRNAnetsim.html.

MiR-451a suppressing BAP31 can inhibit proliferation and increase apoptosis through inducing ER stress in colorectal cancer

The global morbidity and mortality of colorectal cancer (CRC) are ranked the third among gastrointestinal tumors in the world. MiR-451a is associated with several types of cancer, including CRC. However, the roles and mechanisms of miR-451a in CRC have not been elucidated. BAP31 is a predicted target gene of miR-451a in our suppression subtractive hybridization library. Its relationship with miR-451a and function in CRC are unclear. We hypothesized that miR-451a could induce apoptosis through suppressing BAP31 in CRC. Immunohistochemistry and real-time PCR were used to measure BAP31 expressions in CRC tissues and pericarcinous tissues from 57 CRC patients and CRC cell lines. Dual-luciferase reporter assay was used to detect the binding of miR-451a to BAP31. The expression of BAP31 protein in CRC tissues was significantly higher than that in pericarcinous tissues, which was correlated with distant metastasis and advanced clinical stages of CRC patients. The expression of BAP31 was higher in HCT116, HT29, SW620, and DLD cells than that in the normal colonic epithelial cell line NCM460. The expression of BAP31 was absolutely down-regulated when over-expressing miR-451a in HCT116 and SW620 cells compared with control cells. Mir-451a inhibited the expression of BAP31 by binding to its 5'-UTR. Over-expressing miR-451a or silencing BAP31 suppressed the proliferation and apoptosis of CRC cells by increasing the expressions of endoplasmic reticulum stress (ERS)-associated proteins, including GRP78/BIP, BAX, and PERK/elF2α/ATF4/CHOP, which resulted in increased ERS, cytoplasmic calcium ion flowing, and apoptosis of CRC cells. These changes resulting from over-expressing miR-451a were reversed by over-expressing BAP31 with mutated miR-451a-binding sites. Over-expressing miR-451a or silencing BAP31 inhibited tumor growth by inducing ERS. The present study demonstrated that miR-451a can inhibit proliferation and increase apoptosis through inducing ERS by binding to the 5'-UTR of BAP31 in CRC.

Downregulation of microRNA‑30a in bronchoalveolar lavage fluid from idiopathic pulmonary fibrosis patients

MicroRNAs (miRs) are short, highly conserved small noncoding RNA molecules with fundamental roles in regulating gene expression. To identify miR biomarkers associated with idiopathic pulmonary fibrosis (IPF), the expression pattern of miRs in exosomes from bronchoalveolar lavage fluid (BALF) of elderly patients with IPF were evaluated. High‑throughput quantitative detection of miR expression using a microarray indicated that miR‑125b, miR‑128, miR‑21, miR‑100, miR‑140‑3p and miR‑374b were upregulated in patients with IPF, while let‑7d, miR‑103, miR‑26 and miR‑30a‑5p were downregulated. The expression level of miR‑30a‑5p was further examined, and its potential target genes were predicted using target gene prediction analysis software. A direct regulatory association was confirmed between miR‑30a‑5p and TGF‑β activated kinase 1/MAP3K7 binding protein 3 (TAB3) via a dual‑luciferase reporter assay. Overexpression of miR‑30a‑5p decreased TAB3, α‑smooth muscle actin and fibronectin expression in A549 cells with or without transforming growth factor‑β1 treatment. The decreased expression of miR‑30a in the BALF of patients with IPF, along with the consequential increase in TAB3 expression, may be a crucial factor in IPF progression.

A Distributed Classifier for MicroRNA Target Prediction with Validation Through TCGA Expression Data

BACKGROUND

MicroRNAs (miRNAs) are approximately 22-nucleotide long regulatory RNA that mediate RNA interference by binding to cognate mRNA target regions. Here, we present a distributed kernel SVM-based binary classification scheme to predict miRNA targets. It captures the spatial profile of miRNA-mRNA interactions via smooth B-spline curves. This is accomplished separately for various input features, such as thermodynamic and sequence-based features. Further, we use a principled approach to uniformly model both canonical and non-canonical seed matches, using a novel seed enrichment metric. Finally, we verify our miRNA-mRNA pairings using an Elastic Net-based regression model on TCGA expression data for four cancer types to estimate the miRNAs that together regulate any given mRNA.

RESULTS

We present a suite of algorithms for miRNA target prediction, under the banner Avishkar, with superior prediction performance over the competition. Specifically, our final kernel SVM model, with an Apache Spark backend, achieves an average true positive rate (TPR) of more than 75 percent, when keeping the false positive rate of 20 percent, for non-canonical human miRNA target sites. This is an improvement of over 150 percent in the TPR for non-canonical sites, over the best-in-class algorithm. We are able to achieve such superior performance by representing the thermodynamic and sequence profiles of miRNA-mRNA interaction as curves, devising a novel seed enrichment metric, and learning an ensemble of miRNA family-specific kernel SVM classifiers. We provide an easy-to-use system for large-scale interactive analysis and prediction of miRNA targets. All operations in our system, namely candidate set generation, feature generation and transformation, training, prediction, and computing performance metrics are fully distributed and are scalable.

CONCLUSIONS

We have developed an efficient SVM-based model for miRNA target prediction using recent CLIP-seq data, demonstrating superior performance, evaluated using ROC curves for different species (human or mouse), or different target types (canonical or non-canonical). We analyzed the agreement between the target pairings using CLIP-seq data and using expression data from four cancer types. To the best of our knowledge, we provide the first distributed framework for miRNA target prediction based on Apache Hadoop and Spark.

AVAILABILITY

All source code and sample data are publicly available at https://bitbucket.org/cellsandmachines/avishkar. Our scalable implementation of kernel SVM using Apache Spark, which can be used to solve large-scale non-linear binary classification problems, is available at https://bitbucket.org/cellsandmachines/kernelsvmspark.

Predicting Functional MicroRNA-mRNA Interactions

MicroRNAs (miRNAs) are small RNA molecules that play key regulatory roles in general biological processes and disease pathogenesis. These small RNA molecules interact with their target mRNAs to induce mRNA degradation and/or inhibit the translation of mRNAs into proteins. Therefore, identifying miRNA targets is an essential step to fully understand the regulatory effects of miRNAs. Here, we describe a regularized regression approach that integrates the sequence information with the miRNA and mRNA expression profiles for detecting miRNA targets. This method takes into account the full spectrum of gene sequence features of miRNA targets, including the thermodynamic stability, the accessibility energy, and the context features of the target sites,. Given these sequence features for each putative miRNA-mRNA interaction and their expression values, this model is able to quantify the down-regulation effect of each miRNA on its targets while simultaneously estimating the contribution of each sequence feature for predicting functional miRNA-mRNA interactions.

MicroRNA target prediction using thermodynamic and sequence curves

BACKGROUND

MicroRNAs (miRNAs) are small regulatory RNA that mediate RNA interference by binding to various mRNA target regions. There have been several computational methods for the identification of target mRNAs for miRNAs. However, these have considered all contributory features as scalar representations, primarily, as thermodynamic or sequence-based features. Further, a majority of these methods solely target canonical sites, which are sites with "seed" complementarity. Here, we present a machine-learning classification scheme, titled Avishkar, which captures the spatial profile of miRNA-mRNA interactions via smooth B-spline curves, separately for various input features, such as thermodynamic and sequence features. Further, we use a principled approach to uniformly model canonical and non-canonical seed matches, using a novel seed enrichment metric.

RESULTS

We demonstrate that large number of seed-match patterns have high enrichment values, conserved across species, and that majority of miRNA binding sites involve non-canonical matches, corroborating recent findings. Using spatial curves and popular categorical features, such as target site length and location, we train a linear SVM model, utilizing experimental CLIP-seq data. Our model significantly outperforms all established methods, for both canonical and non-canonical sites. We achieve this while using a much larger candidate miRNA-mRNA interaction set than prior work.

CONCLUSIONS

We have developed an efficient SVM-based model for miRNA target prediction using recent CLIP-seq data, demonstrating superior performance, evaluated using ROC curves, specifically about 20% better than the state-of-the-art, for different species (human or mouse), or different target types (canonical or non-canonical). To the best of our knowledge we provide the first distributed framework for microRNA target prediction based on Apache Hadoop and Spark.

AVAILABILITY

All source code and data is publicly available at https://bitbucket.org/cellsandmachines/avishkar.

Integrating full spectrum of sequence features into predicting functional microRNA-mRNA interactions

MOTIVATION

MicroRNAs (miRNAs) play important roles in general biological processes and diseases pathogenesis. Identifying miRNA target genes is an essential step to fully understand the regulatory effects of miRNAs. Many computational methods based on the sequence complementary rules and the miRNA and mRNA expression profiles have been developed for this purpose. It is noted that there have been many sequence features of miRNA targets available, including the context features of the target sites, the thermodynamic stability and the accessibility energy for miRNA-mRNA interaction. However, most of current computational methods that combine sequence and expression information do not effectively integrate full spectrum of these features; instead, they perceive putative miRNA-mRNA interactions from sequence-based prediction as equally meaningful. Therefore, these sequence features have not been fully utilized for improving miRNA target prediction.

RESULTS

We propose a novel regularized regression approach that is based on the adaptive Lasso procedure for detecting functional miRNA-mRNA interactions. Our method fully takes into account the gene sequence features and the miRNA and mRNA expression profiles. Given a set of sequence features for each putative miRNA-mRNA interaction and their expression values, our model quantifies the down-regulation effect of each miRNA on its targets while simultaneously estimating the contribution of each sequence feature to predicting functional miRNA-mRNA interactions. By applying our model to the expression datasets from two cancer studies, we have demonstrated our prediction results have achieved better sensitivity and specificity and are more biologically meaningful compared with those based on other methods.

AVAILABILITY AND IMPLEMENTATION

The source code is available at: http://nba.uth.tmc.edu/homepage/liu/miRNALasso.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

CONTACT

Yin.Liu@uth.tmc.edu.

A Bayesian Framework to Improve MicroRNA Target Prediction by Incorporating External Information

MicroRNAs (miRNAs) are small regulatory RNAs that play key gene-regulatory roles in diverse biological processes, particularly in cancer development. Therefore, inferring miRNA targets is an essential step to fully understanding the functional properties of miRNA actions in regulating tumorigenesis. Bayesian linear regression modeling has been proposed for identifying the interactions between miRNAs and mRNAs on the basis of the integrated sequence information and matched miRNA and mRNA expression data; however, this approach does not use the full spectrum of available features of putative miRNA targets. In this study, we integrated four important sequence and structural features of miRNA targeting with paired miRNA and mRNA expression data to improve miRNA-target prediction in a Bayesian framework. We have applied this approach to a gene-expression study of liver cancer patients and examined the posterior probability of each miRNA-mRNA interaction being functional in the development of liver cancer. Our method achieved better performance, in terms of the number of true targets identified, than did other methods.