MicroRNA (miRNA)
MicroRNAs (miRNAs) are concise (~22 nucleotides) endogenous non-coding RNAs synthesized by RNA polymerase class II enzymes in the nucleus. They play a pivotal role in regulating mRNA expression through complementary base pairing with the 3'-untranslated regions (3' UTRs) of target mRNAs.
Given the voluminous data arising from past mRNA and miRNA studies, computational methods have become indispensable for offering experimental validation with statistically significant outcomes. These computational approaches are recognized as reliable tools for predicting miRNA targets.
Targets of miRNAs
MiRNAs govern mRNA expression by engaging in binding interactions with the 3' UTR of target mRNAs, employing two distinct binding modes:
In the first category, the miRNA's 5' end, spanning two to seven nucleotides, forms a precise Watson-Crick pairing with the 3' UTR of the mRNA. These nucleotide sequences, known as seed regions, are classified as 6mer, 7mer, or 8mer based on the number of involved nucleotides.
Another class of miRNAs regulates mRNA expression by establishing some base pairing in the seed region. To compensate for insufficient binding, additional base pairing occurs on the 3' side of the miRNA.
It is noteworthy that a single mRNA may be under the regulation of multiple miRNAs, and reciprocally, one miRNA may influence the expression of other miRNAs, adding complexity to the regulatory mechanism.
CD Genomics uses authoritative algorithms and sequencing platforms to systematically classify and annotate miRNA, siRNA, piRNA, and unknown small RNA, and then fully explore the regulatory functions of small RNA through base editing analysis, expression level analysis, and target gene prediction.
CD Genomics Small RNA Data Analysis Pipeline
How to Predict the Target of miRNA?
miRNA target prediction involves identifying complementary sequences typically found in the 3' UTR of mRNA, where miRNAs act to hinder translation and suppress protein synthesis. Computational methods play a vital role in predicting the likelihood and specificity of miRNA-mRNA binding. The development of miRNA target prediction tools is grounded in key features, serving as fundamental criteria for algorithm formulation. 6 critical features considered in miRNA target prediction.
Complementary Base Pairing
miRNAs and target mRNAs exhibit complementary base pairing in their sequences, a pivotal aspect of miRNA-mRNA binding. Typically, a complementary sequence is found between the miRNA and the 3' UTR region of the mRNA.
Free Energy
The assessment of bound miRNA and mRNA stability involves calculating their respective free energy. Lower free energy values generally indicate a more robust binding, providing insights into the strength of the interaction.
Conservativeness
The degree of conservation in miRNA targeting sites across species is a crucial factor. Highly conserved sites, preserved through evolution, are more likely to represent authentic miRNA targets.
Structural Features
Secondary structural features of mRNAs, including stability and accessibility, can influence the location of miRNA binding. Accounting for these features is essential in achieving accurate miRNA target predictions.
Site Polymorphism
Some miRNA targets may exhibit polymorphism, featuring multiple possible binding sites. Considering this variability enhances the precision of miRNA target predictions.
Expression of Homologous Genes
miRNAs typically exert their regulatory functions by reducing the expression of target mRNAs. Therefore, the expression levels of homologous genes play a crucial role in predicting miRNA targets, with highly homologous genes presenting a greater likelihood of becoming miRNA regulatory targets.
MicroRNA Target Prediction Database
miRBase stands as a widely recognized public repository for comprehensive miRNA information. Offering a rich array of data, including published miRNA sequences, annotations, and predicted gene targets, it has recently been updated to version 22.1. This update encompasses 38,589 precursor sequences and 48,885 mature body sequences across 271 species. The database facilitates convenient online queries, enabling users to search for known miRNAs and their associated targets using names or keywords.
The TargetScan database focuses on identifying miRNA target genes in animals, leveraging evolutionarily conserved features within target gene sequences. Specifically, it considers conserved 8mer and 7mer sites matched in the seed region, leading to a low false-positive rate. TargetScan covers humans, mice, rats, cows, dogs, orangutans, rhesus monkeys, opossums, chickens, and frogs. It supports both the prediction of target genes by miRNAs and the reverse prediction of related miRNAs by mRNAs.
Developed by the renowned bioinformatics laboratory led by Segal, PITA predicts miRNA targets based on target-site accessibility and free energy. This database primarily focuses on humans, mice, Drosophila, and worms. Users can predict target genes by miRNAs or forecast miRNA information by inputting mRNA names or IDs. PITA allows users to predict and download target relationship data.
miRTarBase specializes in curating experimentally validated miRNA targets and supplements its search results with literature or methodological references. Users can explore, search, and download data from the database, with queries possible based on miRNA, gene, disease, pathway, and other criteria.
The latest version of starBase compiles over 700 CLIP-seq datasets from 23 species. It features diverse visualization interfaces that leverage degradome and histology experiments to uncover miRNA targets. Beyond elucidating miRNA-mRNA relationships, starBase analyzes interactions among lncRNAs, circRNAs, proteins, and mRNAs, shedding light on ceRNA mechanisms.
miRWalk2.0 represents a substantial improvement over its predecessors, offering the most extensive collection of predictive and experimentally confirmed miRNA target interactions. Through automated text mining, it records experimentally validated miRNA target interactions and incorporates pathway information from related resources such as miRTarBase, PhenomiR2.0, miR2Disease, and HMDD. Additionally, it provides insights into proteins implicated in the miRNA regulatory process.
