Tools for Sequence-Based miRNA Target Prediction: What to Choose?

Abstract

MicroRNAs (miRNAs) are defined as small non-coding RNAs ~22 nt in length. They regulate gene expression at a post-transcriptional level through complementary base pairing with the target mRNA, leading to mRNA degradation and therefore blocking translation. In the last decade, the dysfunction of miRNAs has been related to the development and progression of many diseases. Currently, researchers need a method to identify precisely the miRNA targets, prior to applying experimental approaches that allow a better functional characterization of miRNAs in biological processes and can thus predict their effects. Computational prediction tools provide a rapid method to identify putative miRNA targets. However, since a large number of tools for the prediction of miRNA:mRNA interactions have been developed, all with different algorithms, the biological researcher sometimes does not know which is the best choice for his study and many times does not understand the bioinformatic basis of these tools. This review describes the biological fundamentals of these prediction tools, characterizes the main sequence-based algorithms, and offers some insights into their uses by biologists.

Circular RNA CCDC66 targets DCX to regulate cell proliferation and migration by sponging miR‐488‐3p in Hirschsprung's disease

“…It is clear that circRNAs could sequester miRNAs to control the expression level of the target genes. Hence, Bioinformatic prediction (DIANA, Vlachos & Hatzigeorgiou, , http://diana.imis.athena-innovation.gr/DianaTools/index.php; Targetscan, Riffo‐Campos, Riquelme, & Brebi‐Mieville, , http://www.targetscan.org/vert_71/; and miRDB, Wong & Wang, , http://www.mirdb.org/miRDB/) was applied to look for the probable target genes of miR‐488‐3p. Results showed that DCX was closely related with the miR‐488‐3p in all databases.…”

CircRNA CDR1as knockdown inhibits progression of non‐small‐cell lung cancer by regulating miR‐219a‐5p/SOX5 axis

“…The potential binding sites of miR‐219a‐5p and circCDR1as or SOX5 were predicted by starBase (http://starbase.sysu.edu.cn/index.php) or TargetScan (http://www.targetscan.org/vert_72/). The use and parameters of starBase and TargetScan were shown as previous reports . The psiCHECK‐2 dual‐luciferase vectors (Promega, Madison, WI, USA) were used to generate the wild‐type luciferase reporter vectors (circCDR1as‐WT and SOX5‐WT) and corresponding mutant (circCDR1as‐MUT and SOX5‐MUT) by cloning the sequences of circCDR1as or 3′ untranslated regions (UTR) of SOX5 containing miR‐219a‐5p binding sites.…”

Chikungunya virus modulates the miRNA expression patterns in human synovial fibroblasts

“…The high false‐positive rates of target predictions are the major limitations associated with most of the prediction software. Therefore, only those targets were selected, which had an intersection with at least two major algorithms (Figure ) . The total 139 974 genes were predicted by using target prediction tools and 259 genes were selected out of the total predicted genes for further analysis based on the following criteria: (a) role of genes during virus infection; (b) presence of the gene in at least two databases; and (c) significant scores of the target genes (Table S2).…”