Sigillvm Vniversitatis Hafniensis (The faculty of Science) The Bioinformatics Centre University of Copenhagen

From BINF - Bioinformatics Centre

Contents

Methods in microRNA bioinformatics

This page is supporting the article:

Principles of microRNA gene and target prediction
Morten Lindow and Jan Gorodkin
DNA and Cell Biology. 2007, 26(5): 339-351

Below, you can find extended and hyperlinked versions of Table 1 and 2 from the paper.

Target prediction

Algorithm Matching Energy filter Search space reduction Scoring scheme Downloadable program Submit sequence for online search Database of predictions
TargetScanS Seed variations only None Whole genome alignments All equal No No Yes
miRanda Weighted alignment emphasizing seed RNAfold with linker hack AVID alignment of UTRs ad hoc Yes No Yes (Yes)
PicTar Seedmatch or compensatory >33% of energy for perfect match of miRNA Whole genome alignments Maximum likelihood, combined for all sites meeting tresholds No No Yes
RNAhybrid Integrated match and energy Guarantees to find lowest energy site Not implemented extreme value dist. BLAST-like evalue Yes Yes No
Target Boost Boosted genetic algorithm to create weighted sequence motifs that characterize the probable binding characteristics. None Not implemented Yes No Demo only No
Rna22 Best alignment None 'Target Islands' with high density of pattern 'hits' extreme value dist. BLAST-like evalue No Yes No
Diana-microT Dynamic programming - dinucleotides Hybridisation energy < -20 kcal/mole + other rules must be fulfilled about the complete alignment No Modifyable energy cutoff + hard rules No Yes No
MicroInspector Perfect seed, followed by alignment Hybridization energy from Vienna RNA None implemented Sort by energy No Yes No
miTarget Short seed hard cutoff (adjustable) None implemented SVM that weight a combination of features learned from examples No Yes No

Gene prediction

This is an expanded version of the table appearing in our review-paper. This table also includes publications, where the method used is unnamed and not generally available.

Name Method Description Applied to Downloadable program Webserver Database with results Reference Year
MirScanHybrid: Rule based and using empirical values of parametersRNAfold hairpins compared between organismsWorm, human, mouse, pufferfish-YesPMID:126242572003
"no name given"Filtering C. elegans hairpins and search for conservation to other worms and Drosophila and HumanC. elegansPMID:127698492003
"no name given"PMID:127967792003
"no name given"Comparative genomics including mfold predicted hairpins.C. elegansPMID:127478282003
"no name given"Detection of upstream conserved motif of miRNAsWormPMID:153179712004
"no name given"Genomic location of miRNAsHuman and mousePMID:153649012004
MIRFINDERUse comparative genomics and structural homologyRice and ArabidopsisYes-YesPMID:152720842004
"no name given"Hairpins and sequence conservation.Arabidopsis and ricePMID:153450492004
"no name given"Search simulataneous for microRNA and targets. Organisms specific.ArabidopsisPMID:152009562004
"no name given"Conserved genomic regions of miRNAs (Phylogenetic shadowing)Mammalian genomesPMID:156524782005
PalGradeFiltering RNAfold predicted hairpinsHuman and primate genomes---PMID:159654742005
ProMiRGenetic programming. Learning pre-miRNA features. Human conservation to vertebrates-YesYesPMID:159877892005
"no name given"Search for miRNA clusters human, mouse, rat and virusPMID:157822192005
"no name given"General approach to search for regulatory motifs in promotors and 3'UTR. Human mouse rat dogPMID:157356392005
"no name given"Homology search of both pre-miRNA and mature miRNAHuman and mousePMID:156343322005
ERPIN to construct profiles Profiles search to forAnimal genomesPMID:155096082005
findMiRNAmiRNA patterns from training setArabidopsis and Rice--YesPMID:156320922005
"no name given"Using EST to find siRNA [DO they really find microRNAs?]ArabidopsisPMID:159801472005
Triplet-SVMSVMApproach to distingush between real pre-miRNA hairpins from others.Human applied on plant and virus.Yes--PMID:163816122005
BayesmiRNAfind"Naiive Bayes"As mirScan fold 110 nt genomic sequences then apply a bayes filter trained on XXX worm, vertebrate, plant and virus-Yes-PMID:165432772006
microHARVESTERFind miRNAs homolog to existing ones. BLAST then structural filters.Plants-Yes-PMID:163170732006
RNAmicroSVMUse multiple alignments to predict microRNA hairpinsMammals, urochordates, and nematodes.YesYes-PMID:168734722006
miMatcherSVMIntragenomic matching to simultaneously find miRNAs and their targetsarabidopsis, rice, poplar--YesPMID:161593852006

Useful databases

  • miRBase - 'Official' tracker of miRNA-genes
  • RFAM - Database of RNA families and covariance models describing them.
  • miRNAmap - Alternative interface to miRBase + some of their own predictions and expression data
  • Tarbase - Compilation of some experimentally validated miRNA-target interactions

General software for searching for ncRNA