Table 3.

Summary of the most popular bioinformatics approaches for enhancer identification

Name	Computational method	Highlight	Link	Reference
Heintzman et al.	Clustering and correlation of histone marks profiles	High-recognition performance in HeLa	–	[21]
ChromaSig (*)	Identification of specific histone mark motifs and clustering	The method is sensitive enough to capture patterns characterizing different classes of enhancers.	http://bioinformatics-renlab.ucsd.edu/rentrac/ wiki/ChromaSig	[22]
Rye et al.	Clustering of profiles	The results indicate that selection of relevant TFs may be sufficient to identify regulatory elements	–	[23]
Won et. al.	HMMs	State-of-the-art method suggesting that HMMs are capable of integrating information from multiple histone marks for predicting regulatory elements	http://http/nash.ucsd.edu/chromatin.tar.gz	[54]
Boyle et al.	Combination of DHS with TFBSs	Active enhancers usually overlap with open chromatin regions, but not all of the DNA accessible regions correspond to enhancers	–	[25]
ChromHMM (*)	HMMs	State-of-the-art genome annotation method by ENCODE	http://compbio.mit.edu/ChromHMM	[30]
Segway (*)	DBNs	State-of-the-art genome annotation method by ENCODE	http://www.pmgenomics.ca/hoffmanlab/ proj/segway/	[31]
ChroModule	HMMs	Annotated human genome for eight cell lines and improved the AUC compared with state-of-the-art HMM based methods	–	[32]
CSI-ANN (*)	ANNs	Effective combination of ANNs with FDA for FS	http://www.healthcare.uiowa.edu/labs/ tan/CSIANNWebpage.html	[33]
ChromaGenSVM (*)	SVMs	Effective combination of SVMs with GA for optimization and FS	http://sysimm.ifrec.osaka-u.ac.jp/download/ Diego/	[34]
EnhancerFinder	MKL	Functional genomics combined with sequence motifs can accurately identify developmental enhancers	–	[35]
RFECS (*)	RFs	Method less prone to overfitting, which introduces additional novelties on the way enhancer predictions are validated	http://enhancer.ucsd.edu/renlab/RFECS_ enhancer_prediction/	[37]
DEEP (*)	SVMs and ANNs	Novel ensemble-learning-based algorithm with good generalization capabilities in unknown cell lines..	http://cbrc.kaust.edu.sa/deep/	[36]
kmer-SVM (*)	SVMs	Study extensively the enhancer sequence context	http://kmersvm.beerlab.org/	[39]
dREG (*)	SVR	Usage of GRO-seq data combined with regression analysis	https://github.com/Danko-Lab/dREG/	[41]
DELTA (*)	AdaBoost	Introduces the concept of shape features from ChiP-seq data	https://github.com/drlu/delta	[38]
Andersson et al. (*)	eRNA expression analysis	Introduces one of the most accurate features for enhancer identification	http://enhancer.binf.ku.dk/enhancers.php	[40]
CoSBI (*)	Bi-clustering	Reports combination of histone marks with high discriminative power for the category of enhancers	http://www.healthcare.uiowa.edu/labs/tan/ CoSBIWebpage.html	[24]

Note. With (*), are marked the methods that provide source codes or executable files.