Predictive Methods Description.

MiRGate contains in-house predictions produced by the following methods: miRanda, Pita, RNAHybrid, Microtar and TargetScan. You can find more information below.


Betel, D., Koppal, A., Agius, P., et al. (2010).Comprehensive modeling of microRNA targets predicts functional non-conserved and non-canonical sites.Genome biology, 11, R90.
The Miranda algorithm version 3.3a uses dynamic programming to score alignments based of the complementarity of nucleotides, allowing G-U wobble pairs. Thermodynamic stability (energy > 120 kcal) and conservation are used as threshold.


Kertesz, M., Iovino, N., Unnerstall, U., et al. (2007) The role of site accessibility in microRNA target recognition. Nature genetics, 39, 1278-1284.
Probability of interaction by target accessibility (Pita) identifies initial full complementary seeds for each miRNA in the mRNA and computes the free energy of the unbound and bound double strand. It uses a phylogenetic hidden Markov model (Siepel, A., Bejerano, G., Pedersen, J.S., et al. (2005)) called Phastcons; to filter out less conserved predicted target sites. Target sites are filtered with a score of 0.5, which roughly corresponds to conservation across all mammals.


Kruger, J., Rehmsmeier, M. (2006). RNAhybrid: microRNA target prediction easy, fast and flexible. Nucleic acids research, 34, W451-454.
The program finds energetically most favorable hybridization sites avoiding intramolecular hybridization. Poisson approximation of multiple binding sites and calculation of effective numbers of orthologous targets in comparative studies of multiple organisms are assessed. Version 2.2.


Thadani, R., Tammi, M.T. (2006). MicroTar: predicting microRNA targets from RNA duplexes. BMC bioinformatics, 7 Suppl 5, S20.
It is a program based on mRNA sequence complementarity and RNA duplex energy prediction by using Vienna package, assessing the impact of microRNA binding on complete mRNA molecules.


Friedman, R.C., Farh, K.K., Burge, C.B., et al. (2009) Most mammalian mRNAs are conserved targets of microRNAs. Genome research, 19, 92-105
TargetScan place importance on the conservation of the miRNA seed, so that we used mammals EnsEMBL alignments to find out if a target is conserved between species. This algorithm requires perfect seed pairing to score the predictions according the type of the seed match, local AU contribution and mRNA binding site localization.

External Data Description

miRGate contains external information from databases storing curated information regarding experimentally validated targets.
Experimentally validated targets, although are expensive and time consuming are needed to understand the implication of predicted targets and provide valuable information to distinguish weak predictions. In that sense, to increase our understanding in miRNA-UTR targets, miRGate store information coming from several validation methodologies stored in three different databases:


TarBase 6.0 hosts detailed information for each miRNA-gene interaction, ranging from miRNA and gene-related facts to information specific to their interaction, the experimental validation methodologies and their outcomes. All database entries are enriched with function-related data, as well as general information derived from external databases such as UniProt, Ensembl and RefSeq. TarBase hosts 65000 targets manually curated experimentally validated miRNA-gene interactions (Vergoulis, T., Vlachos, I.S., Alexiou, P., et al. (2012). TarBase 6.0: capturing the exponential growth of miRNA targets with experimental support. Nucleic acids research, 40, D222-229.).


MirTarbase 4.5 contains more than 51000 validated miRNA- gene interactions which are collected by manually surveying pertinent literature after data mining of the text systematically to filter research articles related to functional studies of miRNAs (Hsu, S.D., Tseng, Y.T., Shrestha, S., et al. (2014) miRTarBase update 2014: an information resource for experimentally validated miRNA-target interactions. Nucleic acids research, 42, D78-85.).


The last version of the validated target database, includes 2705 records of validated miRNA-target interactions between 644 miRNAs and 1901 target genes in nine animal species (Xiao, F., Zuo, Z., Cai, G., et al. (2009) miRecords: an integrated resource for microRNA-target interactions. Nucleic acids research, 37, D105-110.).


oncomiRDB database aiming at annotating the experimentally verified oncogenic and tumor-suppressive miRNAs from literature. This database only collects items having direct functional evidences: 1) the miRNA regulates at least one cancer-related phenotype or cellular process (such as proliferation, apoptosis, migration and invasion, senescence and cell cycle regulation); or 2) the miRNA directly regulates at least one oncogenic or tumor-suppressive gene verified by luciferase reporter assay. Wang D, Gu J, et al..

1) How do you define a known/unknown UTR?

When an EnsEMBL isoform is not known, we create a predicted UTR using the mode value. The mode value is the value which is repeated the most. So we calculate the mode value between known UTR size in human (142.02bp), mouse (130.92bp) and rat (121.54bp). As the mean value is ~ 130bp. We take as predicted UTR the 130bp after the last exon.

2) Where can I find more information about HAVANA biotypes?

The can be found in the following link :

3) What version of Genecode are you using?

For ENCODE isoform annotation, we use APPRIS which is a system that deploys a range of computational methods to provide value to the annotations of the human genome. APPRIS also selects one of the CDS for each gene as the principal isoform. It defines the principal variant by combining protein structural and functional information and information from the conservation of related species. Rodriguez JM, Maietta P, Ezkurdia I, Pietrelli A, Wesselink JJ, Lopez G, Valencia A, Tress ML. (2013). APPRIS: annotation of principal and alternative splice isoforms. Nucleic Acids Res. 2013 Jan;41(Database issue):D110-7.

4) What is the grouping threshold?

miRGate provides a unique feature among other initiatives, considering an overlap when the binding event between the miRNA seed and the mRNA 3′ UTR occurs in the same genomic position. Hence it is possible to label remarkable agreed predictions when five different algorithms predict the same target (in tense of target site type) in the same RNA coordinates. By default, predictions are colapesd when thy are predicted in the same coordinates +/- 2 nt. But this value can be changed by the user.