Comprehensive comparison of microRNA target prediction methods

in Review Publications December 12, 2014 6,743 Views

MicroRNAs (miRNAs) are short endogenous noncoding RNAs that bind to target mRNAs, usually resulting in degradation and translational repression. Identification of miRNA targets is crucial for deciphering functional roles of the numerous miRNAs that are rapidly generated by sequencing efforts. Computational prediction methods are widely used for high-throughput generation of putative miRNA targets.

Researcher from the University of Alberta review a comprehensive collection of 38 miRNA sequence-based computational target predictors in animals that were developed over the past decade. Their in-depth analysis considers all significant perspectives including the underlying predictive methodologies with focus on how they draw from the mechanistic basis of the miRNA-mRNA interaction. They also discuss ease of use, availability, impact of the considered predictors and the evaluation protocols that were used to assess them. The gene-level evaluation is based on three benchmark data sets that rely on different ways to annotate targets including biochemical assays, microarrays and pSILAC. The researchers offer practical advice on selection of appropriate predictors according to certain properties of miRNA sequences, characteristics of a specific application and desired levels of predictive quality. Finally, they discuss future work related to the design of new models, data quality, improved usability, need for standardized evaluation and ability to predict mRNA expression changes.

Methodologies and the corresponding mechanistic basis of miRNA–mRNA interaction used by the miRNA target predictors

Predictor	Reference	Year published	Model type	Complementarity		Site accessibility		Conservation
Predictor	Reference	Year published	Model type	Seed	Nonseed	Free energy	AU %	Conservation
Stark et al.	[20]	2003	Screening	1–8	miRNA size+5	mFold	✕	a d
TargetScan	[44]	2003	Score	7mer-m8	to 1^st mismatch	Vienna RNA	✓	m r z
DIANA-microT	[54]	2004	Score	✕	38 nt	✓		m
RNAhybrid	[47]	2004	Score	6mer	✕	RNAhybrid	✕	a d
miRanda	[55]	2004	Score	7mer-m8	✕	Vienna RNA	✕	f m r
Rajewsky’s	[23]	2004	Score	1–8	✕	mFold	✕	d
TargetScanS	[56]	2005	Score	6mer	✕	✕	✕	c g h m r
Robins	[57]	2005	Score	2–8	✕	Vienna RNA	✕	✕
Xie et al.	[58]	2005	Score	8mer	✕	✕	✕	g h m r
PicTar	[46]	2005	Score	7mer-A1, 7mer-m8	Remaining	mFold	✕	d
MovingTarget	[59]	2005	Screening	1–8	50 nt	DINAMelt	✕	d
Microlnspector	[60]	2005	Score	7mer-A1, 7mer-m8	✕	Vienna RNA	✕	✕
TargetBoost	[61]	2005	GP	pattern	30 nt	mFold	✕	✕
Stark et al.	[62]	2005	Score	6mer	10^th nt to end	RNAhybrid	✕	d
miTarget	[63]	2006	SVM	2–7	20 nt	Vienna RNA	✓	✕
RNA22	[15]	2006	Score	✕	Pattern	✕	✕	✕
MicroTar	[64]	2006	Score	7mer-A1, 7mer-m8	✕	Vienna RNA	✕	✕
EIMMo	[65]	2007	Bayesian	7mer-A1, 7mer-m8	✕	✕	✕	✓
STarMir	[66]	2007	Score	✕	miRNA size	sFold	✕	✕
PITA	[67]	2007	Score	6mer	✕	Vienna RNA	✕	✕
TargetRank	[68]	2007	Score	6mer	✕	✕	✓	✓
MirTarget2	[43]	2008	SVM	6mer	✕	Vienna RNA	✕	✕
HuMiTar	[48]	2008	Score	6mer	9–13, 14–20 nt	✕	✕	✕
TargetMiner	[49]	2009	SVM	6mer	13–16 nt	✓	✓	✓
TargetSpy	[52]	2010	DS	✕	All	Vienna RNA	✕	✕
Mtar	[53]	2010	ANN	6mer	Remaining	Vienna RNA	✕	✕
mirSVR	[40]	2010	Score	2–7	✕	miRNAbind	✓	✓
SVMicrO	[50]	2010	SVM	5 patterns	Remaining	Vienna RNA	✓	✓
RepTar	[69]	2010	Screening	6mer	Remaining	Vienna RNA	✕	✓
PACMIT	[51]	2011	Screening	✕	Remaining	Vienna RNA	✕	✕
MultiMiTar	[70]	2011	SVM	6mer	13–16 nt	✕	✓	✕
miREE	[71]	2011	SVM	1–8	13–16 nt, remain	Vienna RNA	✓	✕
miRcode	[72]	2012	Screening	7mer-A1, 7mer-m8	✕	✕	✕	P M other V
miRmap	[41]	2012	Regression	6mer	Remaining	Vienna RNA	✓	M
HomoTarget	[73]	2012	ANN	1–8	Remaining	✓	✕	✕
SuperMirTar	[74]	2013	Graph	6mer	12–17 nt	RNAhybrid	✓	✕
Fujiwara’s	[75]	2013	Cis-element	✕	✕	✕	✕	✕
MIRZA	[76]	2013	Bayesian	1–8	Remaining	✓	✕	✕

Fan X, Kurgan L. (2014) Comprehensive overview and assessment of computational prediction of microRNA targets in animals. Brief Bioinform [Epub ahead of print]. [abstract]