UBC Faculty Research and Publications

MIR@NT@N: a framework integrating transcription factors, microRNAs and their targets to identify sub-network… Le Béchec, Antony; Portales-Casamar, Elodie; Vetter, Guillaume; Moes, Michèle; Zindy, Pierre-Joachim; Saumet, Anne; Arenillas, David; Theillet, Charles; Wasserman, Wyeth W; Lecellier, Charles-Henri; Friederich, Evelyne Mar 4, 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12859_2010_Article_4441.pdf [ 1.15MB ]
JSON: 52383-1.0224081.json
JSON-LD: 52383-1.0224081-ld.json
RDF/XML (Pretty): 52383-1.0224081-rdf.xml
RDF/JSON: 52383-1.0224081-rdf.json
Turtle: 52383-1.0224081-turtle.txt
N-Triples: 52383-1.0224081-rdf-ntriples.txt
Original Record: 52383-1.0224081-source.json
Full Text

Full Text

SOFTWARE Open AccessMIR@NT@N: a framework integrating transcriptionfactors, microRNAs and their targets to identifysub-network motifs in a meta-regulation networkmodelAntony Le Béchec1, Elodie Portales-Casamar2, Guillaume Vetter1, Michèle Moes1, Pierre-Joachim Zindy3,Anne Saumet4, David Arenillas2, Charles Theillet4, Wyeth W Wasserman2, Charles-Henri Lecellier5,Evelyne Friederich1*AbstractBackground: To understand biological processes and diseases, it is crucial to unravel the concerted interplay oftranscription factors (TFs), microRNAs (miRNAs) and their targets within regulatory networks and fundamental sub-networks. An integrative computational resource generating a comprehensive view of these regulatory molecularinteractions at a genome-wide scale would be of great interest to biologists, but is not available to date.Results: To identify and analyze molecular interaction networks, we developed MIR@NT@N, an integrativeapproach based on a meta-regulation network model and a large-scale database. MIR@NT@N uses a graph-basedapproach to predict novel molecular actors across multiple regulatory processes (i.e. TFs acting on protein-codingor miRNA genes, or miRNAs acting on messenger RNAs). Exploiting these predictions, the user can generatenetworks and further analyze them to identify sub-networks, including motifs such as feedback and feedforwardloops (FBL and FFL). In addition, networks can be built from lists of molecular actors with an a priori role in a givenbiological process to predict novel and unanticipated interactions. Analyses can be contextualized and filtered byintegrating additional information such as microarray expression data. All results, including generated graphs, canbe visualized, saved and exported into various formats. MIR@NT@N performances have been evaluated usingpublished data and then applied to the regulatory program underlying epithelium to mesenchyme transition(EMT), an evolutionary-conserved process which is implicated in embryonic development and disease.Conclusions: MIR@NT@N is an effective computational approach to identify novel molecular regulations and topredict gene regulatory networks and sub-networks including conserved motifs within a given biological context.Taking advantage of the M@IA environment, MIR@NT@N is a user-friendly web resource freely available athttp://mironton.uni.lu which will be updated on a regular basis.BackgroundThe cells of an organism harbor a common set of geneswhich are differentially regulated in time and spaceby various factors allowing them to adopt distinctphenotypes and to exert various functions. Among theregulators, transcription factors (TFs) and microRNAs(miRNAs) which are small 21-23-nucleotide-long, non-coding RNAs, play a cardinal role in the determinationof cell fate and homeostasis, in physiological and diseaseconditions. While TFs act at the DNA level by bindingto cis-regulatory elements of genes, termed Transcrip-tion Factor Binding Sites (TFBSs) [1-3], miRNAs regu-late gene expression at the post-transcriptional level bybinding to the 3’-untranslated region (3’-UTR) of mes-senger RNAs [4]. They thereby inhibit protein synthesisby triggering the degradation of the target messenger or* Correspondence: evelyne.friederich@uni.lu1Cytoskeleton and Cell Plasticity lab, Life Sciences Research Unit-FSCT,University of Luxembourg, L-1511 Luxembourg, LuxembourgFull list of author information is available at the end of the articleLe Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67© 2011 Le Béchec et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.by inhibiting its translation, contributing to the fine-tuning of gene expression [5,6]. Rather than acting inde-pendently or in parallel, it is now well established thatTFs and miRNAs act in concert in networks to regulatetarget genes in a coordinated manner [7,8]. TFs andmiRNAs are in turn regulated, in part, at transcriptionaland post-transcriptional levels. In line, regulatory nodesmay comprise TFs and miRNAs that form sub-networksincluding fundamental, evolutionary conserved regula-tory motifs such as feedback or feedforward loops(FBL, FFL) [8-12], contributing to the modulation ofgene expression and the adaptation of cells to changesin their environment. For example, these regulatoryschemes play an important role in cell fate determina-tion during embryonic development and during thedifferentiation/dedifferentiation processes of cells, con-ferring them genetic plasticity [13-15].Potentially, a TF binds to the regulatory motifs ofthousands of genes while a miRNA may target severalhundreds of messenger RNAs. Consequently, in silicopredictions of binding sequences of these regulatorsrequire additional filtering to identify those with poten-tial biological relevance. In line, recent studies havedemonstrated that combining binding site predictionswith context-linked, experimental genome-wide co-expression data, is a powerful approach to identify bio-logically meaningful molecular interactions [7,12,16,17].To date, databases and tools have been establishedwhich compile and explore experimentally supportedand predictive data from TF regulations on codinggenes (TF®Gene) [3,18,19], TF regulations on miRNAgenes (TF®miRNA) [20-23] and miRNA regulations onmessenger RNAs (miRNA® gene), [21,24,25]. Whilethese resources and associated tools are useful to predictTF or miRNA binding sites and associated molecularinteractions, an approach which integrates this informa-tion at a genome-scale level to identify miRNA, TF andtarget gene regulatory sub-networks is still not available.Thus, a resource dedicated to the reconstitution ofmeta-regulation networks guided by ‘-omics’ data wouldbe of great interest to users to better understand howthese regulations contribute to biological processes innormal and pathological conditions.Here, we have developed MIR@NT@N (MIRna @NdTranscription factor @nalysis Network), based on agraph-theoretical method to integrate multiple regula-tion levels into a unified model (Figure 1). MIR@NT@Npredicts novel molecular actors and the form of theirinterplay. Based on these predictions or on lists ofknown molecular actors, users can generate regulatorynetworks and extract FBL and FFL sub-networks. Ana-lyses can be contextualized and filtered by associating,for example, large-scale co-expression data. Collectively,MIR@NT@N offers novel applications to gain insightinto the potential mechanisms of action of molecularregulators and their targets, in a given biologicalcontext.ImplementationThe MIR@NT@N applicationThe MIR@NT@N application is an open-access webinterface, which can be accessed as a standalone moduleor through the workflow of M@IA, an environment dedi-cated to integrative biology analyses [26]. MIR@NT@Nis built in the PHP programming language for databasegeneration, data integration, analysis scripts (includinggraph construction and FBL and FFL detection) andinterface. It also uses applications included in M@IA: Rlanguage (http://www.r-project.org) for statistical com-puting and Graphviz tool (http://www.graphviz.org) forinteraction graph generation. Data can be furtherprocessed using any other module of M@IA, suchas automatic gene annotation and data mining basedon ontology or metabolic/signaling pathways. TheMIR@NT@N application is connected to a MySQL rela-tional database integrating information on biologicalJASPAR PAZAR oPOSSUM miRBase MicroCosm microRNA.org Ensembl UniHI TFBS predictions on miRNA PAZAR|JASPAR|miRBase|Ensembl Figure 1 Meta-regulation network model. The MIR@NT@N meta-regulation network model illustrates interactions between threebiological entities, transcription factors (TF), non coding microRNAgenes (miRNA) and coding-genes (Gene). This Gene entityrepresents a target at multiple levels: a DNA sequence (TFregulations), a messenger RNA (miRNA regulations), or a protein(protein-protein interactions). Similarly, edges describe TFregulations (arrows) at DNA level, and miRNA regulations (bluntarrows) at RNA level. Squares represent TFs, diamonds miRNAs andcircles coding-gene targets. The MIR@NT@N database is a large-scale resource which integrates information from multiple availabledatabases: PAZAR, JASPAR and oPOSSUM (for TF regulations),miRBase, MicroCosm and microRNA.org (for miRNA targetpredictions), UniHI (for protein-protein interactions) and Ensembl(for gene annotations). Based on these resources, the MIR@NT@Ndatabase also integrates large-scale information about TFregulations on miRNAs through the prediction of TFBS on upstreamsequences of miRNA precursors.Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 2 of 11entities and their regulations and interactions. TheMIR@NT@N help main page includes an overall descrip-tion of each section, including a quick tutorial and exam-ple files. To guide users in their analysis, MIR@NT@Nalso provides a contextual help available within each sec-tion, by explaining parameters, checking loaded data, andsuggesting analysis refinement.Results and DiscussionThe MIR@NT@N databaseBiological entities were identified and annotated in theMIR@NT@N database using three source databases:PAZAR (Release 2010, January 2010), which includesJASPAR (Release 2009, October 2009) [18,27,28], for TFannotations, including TFBS profiles as Position Fre-quency Matrices (PFMs); miRBase (Release 14, April2010) [25] for the miRNA annotations including locali-zation on genome sequence; and Ensembl (Release 56,October 2009) [29] for TF and coding-gene identifiermapping.To integrate TF regulations on coding-genes(TF®Gene), we combined PAZAR [28] which providespublic TF regulatory data, and oPOSSUM (Release 2.0,January 2007) [19], a large scale database which amongother features, predicts TFBSs conserved between spe-cies, using TFBS profiles from the JASPAR database.Further, we extracted from oPOSSUM all TF®Generegulations predicted in the 10 kb upstream and 5 kbdownstream region of genes, with a score threshold of0.85, and a high conservation level (top percentile of0.010 and minimum identity of 80%). For each of theJASPAR profiles, we calculated the correspondence ofthe scores with empirically derived p-values for a com-mon reference DNA sequence (see “Motif Scoring Pro-cedure and Computation of JASPAR Profile MatrixScore p-values” section on MIR@NT@N website formore details) and established that, for 97% (127 of 130)of the binding site profiles, the applied 0.85 thresholdcorresponds to a p-value no more permissive than p <0.01. Present databases [21-23] do not provide sufficientinformation about TFBSs within genes encoding miR-NAs (TF®miRNA) required for building a large-scalemeta-regulation model. TransmiR provides a limitednumber of experimentally validated regulations for mul-tiple species [20,22]. MiRGen offers the downloading oflarge-scale predicted regulations, but only for Humanand Mouse, and without TFBS scores and locations [21],whereas PuTmiR provides scores only for Human [23].Regulation of transcription of coding and miRNA geneshas been proposed to be similar. This is based on theobservation that promoter regions of both share com-mon features such as the presence of CpG islands andspecific histone modification markers [30]. In furthersupport of common regulatory mechanisms, it has beenshown that a same transcription factor can regulateboth, protein-encoding and miRNA genes [31]. Thus,we have used a standard TFBS detection algorithm [3]and TFBS profiles from the JASPAR database to predictTF®miRNA regulations on a large scale. PFMs wereconverted into Position Weight Matrices (PWMs) andused to predict potential TFBSs in 10 kb sequenceslocated upstream of miRNA precursors, extracted fromEnsembl database, according to pre-miRNA localizationprovided by miRBase. To limit the noise of false predic-tions, only predicted TFBSs with a score higher than0.65 were integrated into MIR@NT@N database.To refine the TFBS prediction on the miRNAupstream sequences, we provide additional informationon TFBS location within “CpG islands” (CGI), regionswhich are frequently associated with promoter regions[30,32]. CGI were predicted (for Human, Mouse andRat) with CpGcluster [33], a distance-based CGI-finderalgorithm, and CpGProd [34], a tool that identifies pro-moter regions associated with CGI.To integrate miRNA-dependent regulations (miRNA®-gene) into MIR@NT@N database, we combined the miR-Base Targets database, rebranded as MicroCosm (Release5, September 2009) and hosted at the EBI (release 5), andmicroRNA.org (Release September 2008) [35]. Eachresource can be used, through the MIR@NT@N applica-tion, separately with scores (from the minimum score of13 to maximum score of 23 for MicroCosm, and from theminimum score of 140 to the maximum score of 205 formicroRNA.org) derived from the miRanda algorithm(John et al., 2004), or simultaneously with a unified score(derived by a non-linear transformation and distributeduniformly between 0 and 1).In addition, we integrated protein-protein interactionsfrom the UniHI database [36], motivated by the ideathat clustered miRNAs can coordinately regulate pro-tein-protein interaction networks [37].Thereby, for 7 species (Caenorhabditis elegans, Daniorerio, Drosophila melanogaster, Gallus gallus, Homosapiens, Mus musculus and Rattus norvegicus),MIR@NT@N database contains 3 638 miRNAs, 335TFs, 68 202 coding-genes as well as a large number ofpredicted interactions for a common standard scorethreshold of 0.85 (211 783 miRNA®Gene, 32 224TF®miRNA and 273 264 TF®Gene).The MIR@NT@N database is publically available onthe website, which proposes 1) a dump file of the data-base in a SQL format, 2) a file (tab-delimiter format)with all TFBS scores calculated from miRNA upstreamsequences and TF profiles from PAZAR, and 3) a file(tab-delimiter format) of the meta-regulation network,combining all regulations (TF®miRNA, miRNA®Geneand TF®Gene) for a common standard score thresholdof 0.85.Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 3 of 11Overview on MIR@NT@NThe MIR@NT@N application works within a meta-regu-lation network model (Figure 1) in order to a) identifynovel major regulators and targets based on an input listof actors, through interaction graph analysis and sub-net-work detection; and b) construct networks with well-defined actors with a presumed role in a given context.Thus, two types of queries are involved. The first typeallows searching for novel key actors in a biological con-text, using TF/gene/miRNA lists as input (including quan-titative expression profiles generated by transcriptomics/proteomics experiments). This query includes three sec-tions: (i) “Transcription Factor regulation” which statisti-cally predicts potential TFs regulating a list of miRNAs, orconversely miRNAs regulated by a list of TFs; (ii) “miRNAregulation” which statistically predicts the significant tar-gets of a list of miRNAs or the miRNAs targeting a list ofgenes; and (iii) “Regulation Network” which combinesboth TF and miRNA regulation predictions to reconstitutemeta-regulation networks and allows detection of regula-tory motifs such as FBL or FFL. The second type of queryprovides an overview on any TF, gene or miRNA, includ-ing their interactions: The “Quick Search” rapidly retrievesinformation on any actor, its regulators and/or targets,while the “Quick Network” generates regulation networksfrom a list of actors presumed to be involved in a particu-lar biological context, and also allows the extraction ofsub-networks including regulatory motifs.As described below, the performance of MIR@NT@Nwas evaluated with published, experimentally validateddata and further highlighted in a biological case studyon epithelium to mesenchyme transition (EMT). EMT isan evolutionary conserved biological process involvingthe reprogramming of regulatory networks, includingTFs, miRNAs and their targets, in epithelial cells duringgastrulation, neural crest cell migration in embryogen-esis. In adults EMT is reactivated in pathological situa-tions such as wound healing, carcinoma progression,and fibrosis [14,38,39].Transcription Factor regulationThis section reports potential TF®miRNA regulationgiven a list of TFs or miRNAs to identify novel TFregulators and miRNA targets. The result is a table ofTFs or miRNAs, filtered and ranked by their relevanceaccording to several criteria (Figure 2A) including: theA B C Figure 2 Output of “Transcription Factor regulation”. (A) Result of the query predicting TFs which potentially regulate the miR-200 family(cluster of down-regulated hsa-mir-200a, hsa-mir-200b, hsa-mir-429 represented here as down-regulated, green color code) with stringentcriteria: quality score ≥ 0.85, number of miRNA by TF ≥ 3, and Fisher test p-value ≤ 0.05. The four predicted TFs are shown. Results are sorted bythe quality score (lower panel). Information on all TF scores, including targeted miRNAs and the number of potential TFBS in a frame are given(shown only for ZEB2). All results, including generated graphs, can be visualized, memorized and exported in various formats. (B) Generatedregulation graph with all input miRNAs shown here as down-regulated (diamonds in green) and predicted TFs (squares in gray). Edges representregulations, and the gray canonical color code corresponds to the quality score. (C) Detail result interface showing the hsa-mir-200b upstreamsequence (black line) with all TFBS (black boxes) predicted for ZEB2 and predicted promoter sequences (yellow or orange bar below the blackline). A table ranks TFBS by quality score, and includes: “ID” (corresponding to the position in the sequence), sequence size, binding sequence,quality score, localization (within the upstream sequence and the genome through a hypertext link), and potential localization within a predictedpromoter (orange boxes).Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 4 of 11quality score, the Fisher test p-value, the number ofTFBSs and TF®miRNA regulations. Results can bevisualized through an interaction graph (Figure 2B)using a gray scale canonical color code to convey pre-diction scores. To facilitate the detection of regulatoryclusters, the graph includes expression information (ifprovided as input) using the green/red canonical colorcode. In addition, to identify clusters of miRNAsregulated by the same TFs, or clusters of TFs whichregulate the same miRNAs, an analysis of the interac-tion graph provides “square” and “curvature” graphs[26]. To refine the prediction analysis, all correspond-ing TFBSs can be visualized through a user-friendlyinterface (Figure 2C) which provides the bindingsequence, its length, the quality score, the localizationon the miRNA upstream sequence and in the genome(with a link to Ensembl), and information about pre-dicted promoters using a canonical color code forprediction scores (from yellow to red). All results(tables and graphs) provide links to external knowl-edge sources (PAZAR for TFs, miRBase for miRNAs,Ensembl for genes and TFBS localization). Results canbe exported and stored for further analysis, using forinstance the M@IA environment [26] or externalapplications.To illustrate the performances of these functions, weidentified TFs predicted to regulate the miR-200 family,including miR-200a, miR-200b and miR-429, which areimportant for the maintenance of the epithelial pheno-type and in the prevention of EMT [40]. Using stringentcriteria we identified four TFs (Figure 2AB) includingZEB2 which has recently been reported to directly inter-act with E-boxes of the miR-200 promoter [14]. Thepredicted TFBSs of ZEB2 can be located on the miR-200 promoter by clicking on the ZEB2 table, yielding 1to 9 sites with the criteria 0.9 and 0.65, respectively(Figure 2C). Interestingly, one of the predicted TFBSs islocated within the experimentally identified region ofthe miR-200 promoter (Bracken et al., 2008) shown tobe negatively regulated by the related transcription fac-tor ZEB1, mediated through paired E-boxes.miRNA regulationThis section determines potential miRNA®Gene regula-tions from a list of miRNAs or other genes to identifynovel actors, i.e. miRNA regulators and targeted genes.The result is a table of genes filtered and ranked bytheir relevance to the input list of miRNA, using alter-native criteria (Figure 3A): MicroCosm and/or micro-RNA.org scores (or corresponding unified score), FisherA B Figure 3 Output of “miRNA regulation”. (A) Result of the prediction of target genes for the miRNA 200 family (cluster of down-regulated hsa-mir-200a, hsa-mir-200b, hsa-mir-429) with medium stringency criteria (miRBase score ≥ 16 and p-value ≤ 0.05, microRNA.org score ≥ 150, at leasttwo miRNAs per gene). The output list was filtered with a list of 132 up-regulated genes. The 20 predicted target genes presented in the table(lower panel) were sorted by the Fisher test p-value. Information can be obtained for each target gene, including its miRNA regulators andunified prediction scores (panels for the first gene is shown). (B) Generated regulation graph with all input miRNAs shown as down-regulated(diamonds in green) and predicted target genes shown as up-regulated (ellipses and square in red). Edges represent regulations, and the graycanonical color code corresponds to the quality score.Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 5 of 11test p-values, number of targets per miRNA and thenumber of targeted sequences (boxes) per gene. Inver-sely, this section can provide, using the same para-meters, a list of miRNAs predicted to regulate a givenlist of genes.As described above in the “TF regulation” section,results can also be filtered using a specific list of data tocontextualize the study, visualized through the sametype of interaction graphs (Figure 3B). We illustratedthis feature by predicting genes that are potentially tar-geted by three miR-200 family members. MIR@NT@Npredicted 934 genes to be at least targeted by two miR-200 family members, using the criteria described inlegend of Figure 3. As these miRNAs are known to bedown-regulated in EMT [40], we contextualized thestudy with a biological filter using a list of 132 genesfound to be up-regulated in experimentally inducedEMT [17,41], reasoning that messengers with negativelycorrelated expression levels may be targets of the miR-200 family [17,41]. Twenty genes were predicted to betargeted by miR-200 family members (Figure 3A and3B). The list included FN1, an experimentally validatedtarget of miR-200 [42], genes reported to play an impor-tant role in EMT [43,44] as well as genes with so far nodescribed role in this process, yielding valuable hypoth-eses for experimental investigations.Regulation network generationThis section combines “TF regulation” and “miRNA reg-ulation” interfaces to allow the construction of meta-regulation networks (Figure 4A), with an orientationtowards the detection of network motifs and the identi-fication of multiple target genes, for both TFs and miR-NAs. Within a specific context, the user may identify,from a list of miRNAs, both novel molecular actors andthe nature of the regulation, highlighting fundamentalregulatory motifs [10]. These motifs include FBLs con-sisting in a reciprocal regulation of a TF and a miRNA(Figure 4B), the TF controlling the miRNA and themiRNA regulating the TF [45]. The FBL modulates theactivity of regulators, which is crucial for the spatio-temporal control of their function. On the other hand, aFFL is a regulatory system in which a regulator A regu-lates another regulator B, and both regulators regulate acommon target C [10,11,46]. In MIR@NT@N, FFLs caninvolve a miRNA regulator (FFL-miRNA, Figure 4C) ora TF regulator (FFL-TF, Figure 4D). In addition,MIR@NT@N includes the concept of indirect FFLs(Figure 4E) in which the regulation of the miRNA bythe TF is exerted by an intermediate TF.Users can inform the system by providing a list ofmolecular interactions. For example, the user can use alist of miRNA®gene interactions experimentallyinferred from microarray data combining genes andmiRNA expression or a list of published TF®miRNAinteractions. For this purpose, published and experimen-tally validated TF®miRNA interactions [20,22,23] areprovided and can be used as a filter.To demonstrate regulatory motif detection, we ana-lyzed TF®miRNA regulations from published data byQui et al., including TransmiR data [20,22]. For the 19human TFs found in common within the MIR@NT@Nand Qiu databases, we observed that 81% of the interac-tions listed in the Qiu database were predicted byMIR@NT@N with a TFBS score higher than 0.65, and43% with a TFBS score higher than 0.85 (Additional file1). Using entire MIR@NT@N database, we extractedputative FBLs (Figure 5A) and FFLs (Figure 5B),C D E A B Figure 4 Meta-regulation Network motifs . (A) The meta-regulation network model integrates TFs, miRNAs and target geneswith their regulations. This model allows describing two biologicallyrelevant systems (FBL and FFL) that can be modelled as “networkmotifs”. (B) The FBL “network motif” model describes a reciprocalregulation between a TF and a miRNA. (C, D, E) The FFL “networkmotif” model that includes additional target genes, can beillustrated as three distinct models: the FFL-miRNA model describesthe regulation of a TF and a targeted gene by the same miRNA (C),the FFL-TF model describes the regulation of a miRNA and atargeted gene by the same the TF (D) and the FFL-indirect modelintegrates an additional TF in the FFL-TF (E).Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 6 of 11B A Figure 5 Validation of FBL and FFL motif predictions. 108 published TF®miRNA regulations (Human) were identified from Qiu andcolleagues [22] for the 19 TFs in common with MIR@NT@N database. FBL (A) and FFL motifs B) were identified from the input list of miRNAsincluded in Qiu database, and by filtering actors with the 19 common TFs and with the 108 published TF®miRNA regulations. Non stringentcriteria were used for TF regulation predictions (TFBS score ≥ 0.65, TFBS length ≥ 6, number of miRNA per TF ≥ 1, genes targeted by at least 1TF), and to filter miRNA regulation predictions we used criteria corresponding to a unified score ≥ 0.8 (miRBase score ≥ 17 and p-value ≤ 0.01,and MicroRNA.org score ≥ 152), and 1 gene per miRNA. Squares represent TFs, diamonds represent miRNAs, and edges represent predictedregulations (blue scale color code used for the prediction score). For FBL sub-network, edges represent double regulations.Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 7 of 11including well-documented FBLs implicating an E2F TFfamily member and several miRNA families [47,48], theZEB “Zinc Finger E-Box” TF family and the miR-200family [49,50], and YY1 and hsa-mir-29a [51]. Moreover,we predicted hsa-mir-29a to be regulated by NFKB1 andMYC (Figure 5B), consistent with previous reports[51,52] and only recently identified to be co-regulatorsof their common target mir-29a [53].Collectively, these results underline the efficiency ofMIR@NT@N to generate an overview of a regulatorynetwork and to detect core sub-networks within a biolo-gical system.Quick Search and Quick Network interfacesQuick search and Quick network interfaces allowsearching for regulations between known or assumedactors of a biological context. The “Quick Search” sec-tion is a full text search engine that provides data perti-nent to specific entities (miRNA, TF or target gene).Information about each biological item is availablethrough hypertext links to external data sources(Ensembl for genes, PAZAR for TFs and miRBase formiRNAs). Potential TF/miRNA regulations and pre-dicted TFBSs are accessible through an internalMIR@NT@N application pipeline.The “Quick network” is a powerful application toextract information from a list of TFs, miRNAs andother genes with a presumed function within a biologi-cal context, as supported by literature or experimentaldata. The user can retrieve corresponding regulatorypredictions and generate a network of predicted interac-tions as a comprehensive graph, yielding information onthe interaction mechanisms of the analyzed actors.Functional motifs (FFL and FBL) can be detected toidentify major actors and targets organized into regula-tory sub-networks. The respective quality score thresh-olds of TF and miRNA regulations can be modulatedthrough a cursor and information about protein interac-tions can be integrated into the network (described asexperimentally validated in UniHI). A cross-species net-work analysis is possible by selecting different speciesassociated with the input symbol (e.g. the input symbol“hsa-mir-200a” will be changed into “mmu-mir-200a” ifthe “Mus musculus” species is selected). The output isan exportable interaction graph recapitulating all pre-dicted interactions and which is linked to externalresources (Figure 6).The miR-200 family served as an example to illustratehow the “Quick Network” interface generates regulatorynetworks in a given context (Figure 6). The generatednetwork recapitulates the results described above andintegrates the predictions of the TFBS on coding genes(Figure 6A). The FBL function suggests the presence ofa double negative FBL between mir-200a, mir-200b,mir-429 and ZEB2 (Figure 6B), as is described in the lit-erature [54]. Target genes already described in EMT,such as ROCK2 [44] and TEAD1 [55] were highlightedfrom the FFL network (Figure 6C).Future extensions of MIR@NT@NMIR@NT@N, which takes advantage of the M@IA envir-onment [26], can be readily extended to include additionalmiRNA target prediction databases (such as TargetScan[56] or PicTar [57]) or more TF binding profiles from col-lections that use a standard PFM format. The PWM meth-ods utilized within MIR@NT@N are well-established, butlikely to be replaced with more advanced models in thenear future. High-throughput sequencing coupled to chro-matin immunoprecipitation now routinely generatescollections of ~103 binding sites, providing richer descrip-tions of binding properties of TFs. New algorithms areemerging which build on such data to describe patternsusing higher-order models to account for interactiveeffects between positions. However, the rapidly emergingchanges have not stabilized, so we applied the establishedmethodology within the source database in the oPOSSUMsystem. We intend to upgrade MIR@NT@N when a newmotif scoring procedure is supported by the JASPAR data-base of binding profiles.Moving forward, novel data classes will be implemen-ted into MIR@NT@N, such as histone modifications oralternative splicing that play central roles in geneexpression and for which databases are already available[58,59]. We will incorporate more knowledge sources,such as known promoter sequences and experimentallyvalidated TF-miRNA regulations [20,22,60].ConclusionsHere, we described MIR@NT@N, available as an open-access web application at http://mironton.uni.lu, whichidentifies meta-regulation networks implicating TFs,miRNAs and target genes. The possibility to predict TF-and miRNA-mediated regulations at a genome-widescale is an important novel feature of MIR@NT@N.MIR@NT@N facilitates the analyses of “-omics” data (i.e.any experiment made at a genome scale such as tran-scriptomics and proteomics analyses) and allows detec-tion of relevant molecular interactions and associatedregulatory motifs (e.g. FFL). Users analyzing complexspatio-temporal gene regulation data can obtain experi-ment-suitable insights into the regulatory mechanismsgoverning cellular processes.Availability and requirementsProject name: MIR@NT@NProject home page: http://mironton.uni.luOperating system(s): Platform independentProgramming language: PHP, HTML, Javascript, RLe Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 8 of 11Other requirements: M@IA environment includingApache 1.3 or higher, MySQL 4.0 or higher, R 2.0 orhigher, GraphvizLicense: GNU GPLAny restrictions to use by non-academics: licenceneededAdditional materialAdditional file 1: MIR@NT@N predictions for TF®miRNA regulationsin Qiu et al. Table providing MIR@NT@N database predictions(maximum score, maximum length and number of TFBS) for TF®miRNAregulations described in Qiu et al., for the 19 human TFs found incommon.List of AbbreviationsTF: Transcription Factor; PFM: Position Frequency Matrices; PWM: PositionWeight Matrices; TFBS: Transcription Factor Binding Site; miRNA: microRNA;CGI: CpG Island; FBL: Feedback loop; FFL: Feedforward loop; EMT: Epitheliumto Mesenchyme Transition.AcknowledgementsThis work has been supported by the National Research Fund, Luxembourg(BIOSAN 07/12, AFR fellowship to A.L.B, TR-PDR BFR08-084), the “FondationLuxembourgeoise Contre le Cancer”, the CNRS, MSFHR, CICHR and theFrench Embassy in Canada. A.S. is a recipient of a BFR fellowship (08/046)from the Ministère de la Culture, de l’Enseignement Supérieur et de laRecherche of Luxembourg. W.W.W. was supported as Scholar of the MichaelSmith Foundation. We especially thank E Schaffner-Reckinger, Jean Mullerand Christian Delamarche for critical reading of the manuscript and theirhelpful comments, and Matthias E. Futschik for providing us UniHI database.Author details1Cytoskeleton and Cell Plasticity lab, Life Sciences Research Unit-FSCT,University of Luxembourg, L-1511 Luxembourg, Luxembourg. 2Centre forMolecular Medicine and Therapeutics, Child and Family Research Institute,University of British Columbia, 950 West 28th Avenue, Vancouver, BC V5Z4H4, Canada. 3Structure and Function of the Cell Nucleus, Institute forResearch in Immunology and Cancer (IRIC), Université de Montréal, Montréal(Québec), Canada. 4Institut de Recherche en Cancérologie de MontpellierINSERM U896, Université Montpellier1, CRLC Val d’Aurelle Paul Lamarque,Montpellier, F-34298, France. 5Institut de Génétique Moléculaire deMontpellier UMR 5535 CNRS, 1919 route de Mende, F-34293 Montpelliercedex 5, France.Authors’ contributionsALB designed the MIR@NT@N approach and the software. EP-C and WWWhelped in JASPAR/PAZAR incorporation and TFBS predictions, contributedconstructive suggestions to the study and to the manuscript. C-HL, GV, andMM contributed knowledge on the molecular biology of TF/miRNAregulations and EMT. DA and WWW contributed to the computation ofJASPAR profile matrix score p-values. AS, EF, P-JZ and CT providedA B C Figure 6 Regulatory networks in EMT generated by the Quick Network interface. (A) Networks were generated using the down-regulatedmiR-200 family (diamonds in green), the four TFs (squares in gray) predicted to regulate these miRNAs and the twenty predicted up-regulatedtarget gene list (ellipses in red) predicted to be regulated by these miRNAs. Scores of 0.85 or 0.8 were used for TF regulations or miRNAregulations, respectively. (B) FBL extracted from the previous network (A), showing a regulation loop between ZEB2 and miR-200 family. (C) FFLdetected from the previous network (A), focusing on genes targeted by both, TFs and miRNAs.Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 9 of 11intellectual support and discussed the results. ALB, EF and C-HL wrote themanuscript. All authors read and approved the final manuscript.Received: 27 July 2010 Accepted: 4 March 2011Published: 4 March 2011References1. Pan Y, Tsai CJ, Ma B, Nussinov R: Mechanisms of transcription factorselectivity. Trends Genet 2010, 26(2):75-83.2. Meireles-Filho AC, Stark A: Comparative genomics of gene regulation-conservation and divergence of cis-regulatory information. Curr OpinGenet Dev 2009, 19(6):565-570.3. Wasserman WW, Sandelin A: Applied bioinformatics for the identificationof regulatory elements. Nat Rev Genet 2004, 5(4):276-287.4. Siomi H, Siomi MC: Posttranscriptional regulation of microRNA biogenesisin animals. Mol Cell 2010, 38(3):323-332.5. Pillai RS, Bhattacharyya SN, Artus CG, Zoller T, Cougot N, Basyuk E,Bertrand E, Filipowicz W: Inhibition of translational initiation by Let-7MicroRNA in human cells. Science 2005, 309(5740):1573-1576.6. Fabian MR, Sonenberg N, Filipowicz W: Regulation of mRNA translationand stability by microRNAs. Annu Rev Biochem 2010, 79(2010):351-379.7. Bandyopadhyay S, Bhattacharyya M: Analyzing miRNA co-expressionnetworks to explore TF-miRNA regulation. BMC Bioinformatics 2009,10:163.8. Martinez NJ, Walhout AJ: The interplay between transcription factors andmicroRNAs in genome-scale regulatory networks. Bioessays 2009,31(4):435-445.9. Re A, Cora D, Taverna D, Caselle M: Genome-wide survey of microRNA-transcription factor feed-forward regulatory circuits in human. Mol Biosyst2009, 5(8):854-867.10. Shalgi R, Lieber D, Oren M, Pilpel Y: Global and local architecture of themammalian microRNA-transcription factor regulatory network. PLoSComput Biol 2007, 3(7):e131.11. Tsang J, Zhu J, van Oudenaarden A: MicroRNA-mediated feedback andfeedforward loops are recurrent network motifs in mammals. Mol Cell2007, 26(5):753-767.12. Tsang JS, Ebert MS, van Oudenaarden A: Genome-wide dissection ofmicroRNA functions and cotargeting networks using gene setsignatures. Mol Cell 2010, 38(1):140-153.13. Egli D, Birkhoff G, Eggan K: Mediators of reprogramming: transcriptionfactors and transitions through mitosis. Nat Rev Mol Cell Biol 2008,9(7):505-516.14. Cano A, Nieto MA: Non-coding RNAs take centre stage in epithelial-to-mesenchymal transition. Trends Cell Biol 2008, 18(8):357-359.15. Wang Y, Russell I, Chen C: MicroRNA and stem cell regulation. Curr OpinMol Ther 2009, 11(3):292-298.16. Malhotra D, Portales-Casamar E, Singh A, Srivastava S, Arenillas D, Happel C,Shyr C, Wakabayashi N, Kensler TW, Wasserman WW, et al: Global mappingof binding sites for Nrf2 identifies novel targets in cell survival responsethrough ChIP-Seq profiling and network analysis. Nucleic Acids Res 2010.17. Vetter G, Saumet A, Moes M, Vallar L, Le Bechec A, Laurini C, Sabbah M,Arar K, Theillet C, Lecellier CH, et al: miR-661 expression in SNAI1-inducedepithelial to mesenchymal transition contributes to breast cancer cellinvasion by targeting Nectin-1 and StarD10 messengers. Oncogene 2010.18. Portales-Casamar E, Thongjuea S, Kwon AT, Arenillas D, Zhao X, Valen E,Yusuf D, Lenhard B, Wasserman WW, Sandelin A: JASPAR 2010: the greatlyexpanded open-access database of transcription factor binding profiles.Nucleic Acids Res 2010, , 38 Database: D105-110.19. Ho Sui SJ, Fulton DL, Arenillas DJ, Kwon AT, Wasserman WW: oPOSSUM:integrated tools for analysis of regulatory motif over-representation.Nucleic Acids Res 2007, , 35 Web Server: W245-252.20. Wang J, Lu M, Qiu C, Cui Q: TransmiR: a transcription factor-microRNAregulation database. Nucleic Acids Res 2010, , 38 Database: D119-122.21. Alexiou P, Vergoulis T, Gleditzsch M, Prekas G, Dalamagas T, Megraw M,Grosse I, Sellis T, Hatzigeorgiou AG: miRGen 2.0: a database of microRNAgenomic information and regulation. Nucleic Acids Res 2010, , 38Database: D137-141.22. Qiu C, Wang J, Yao P, Wang E, Cui Q: microRNA evolution in a humantranscription factor and microRNA regulatory network. BMC Syst Biol2010, 4(1):90.23. Bandyopadhyay S, Bhattacharyya M: PuTmiR: a database for extractingneighboring transcription factors of human microRNAs. BMCBioinformatics 2010, 11:190.24. Griffiths-Jones S, Grocock RJ, van Dongen S, Bateman A, Enright AJ:miRBase: microRNA sequences, targets and gene nomenclature. NucleicAcids Res 2006, , 34 Database: D140-144.25. Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ: miRBase: tools formicroRNA genomics. Nucleic Acids Res 2008, , 36 Database: D154-158.26. Le Bechec A, Zindy P, Sierocinski T, Petritis D, Bihouee A, Le Meur N,Leger J, Theret N: M@IA: a modular open-source application formicroarray workflow and integrative datamining. In Silico Biol 2008,8(1):63-69.27. Sandelin A, Wasserman WW: Prediction of nuclear hormone receptorresponse elements. Mol Endocrinol 2005, 19(3):595-606.28. Portales-Casamar E, Arenillas D, Lim J, Swanson MI, Jiang S, McCallum A,Kirov S, Wasserman WW: The PAZAR database of gene regulatoryinformation coupled to the ORCA toolkit for the study of regulatorysequences. Nucleic Acids Res 2009, , 37 Database: D54-60.29. Hubbard TJ, Aken BL, Ayling S, Ballester B, Beal K, Bragin E, Brent S, Chen Y,Clapham P, Clarke L, et al: Ensembl 2009. Nucleic Acids Res 2009, , 37Database: D690-697.30. Marson A, Levine SS, Cole MF, Frampton GM, Brambrink T, Johnstone S,Guenther MG, Johnston WK, Wernig M, Newman J, et al: ConnectingmicroRNA genes to the core transcriptional regulatory circuitry ofembryonic stem cells. Cell 2008, 134(3):521-533.31. Schanen BC, Li X: Transcriptional regulation of mammalian miRNA genes.Genomics 2010.32. Ozsolak F, Poling LL, Wang Z, Liu H, Liu XS, Roeder RG, Zhang X, Song JS,Fisher DE: Chromatin structure analyses identify miRNA promoters. GenesDev 2008, 22(22):3172-3183.33. Hackenberg M, Previti C, Luque-Escamilla PL, Carpena P, Martinez-Aroza J,Oliver JL: CpGcluster: a distance-based algorithm for CpG-islanddetection. BMC Bioinformatics 2006, 7:446.34. Ponger L, Mouchiroud D: CpGProD: identifying CpG islands associatedwith transcription start sites in large genomic mammalian sequences.Bioinformatics 2002, 18(4):631-633.35. Betel D, Wilson M, Gabow A, Marks DS, Sander C: The microRNA.org resource:targets and expression. Nucleic Acids Res 2008, , 36 Database: D149-153.36. Chaurasia G, Malhotra S, Russ J, Schnoegl S, Hanig C, Wanker EE,Futschik ME: UniHI 4: new tools for query, analysis and visualization ofthe human protein-protein interactome. Nucleic Acids Res 2009, , 37Database: D657-660.37. Yuan X, Liu C, Yang P, He S, Liao Q, Kang S, Zhao Y: Clustered microRNAs’coordination in regulating protein-protein interaction network. BMC SystBiol 2009, 3:65.38. Kalluri R, Weinberg RA: The basics of epithelial-mesenchymal transition. JClin Invest 2009, 119(6):1420-1428.39. Thiery JP, Acloque H, Huang RY, Nieto MA: Epithelial-mesenchymaltransitions in development and disease. Cell 2009, 139(5):871-890.40. Gregory PA, Bracken CP, Bert AG, Goodall GJ: MicroRNAs as regulators ofepithelial-mesenchymal transition. Cell Cycle 2008, 7(20):3112-3118.41. Vetter G, Le Bechec A, Muller J, Muller A, Moes M, Yatskou M, Al Tanoury Z,Poch O, Vallar L, Friederich E: Time-resolved analysis of transcriptionalevents during SNAI1-triggered epithelial to mesenchymal transition.Biochem Biophys Res Commun 2009, 385(4):485-491.42. Cochrane DR, Spoelstra NS, Howe EN, Nordeen SK, Richer JK: MicroRNA-200c mitigates invasiveness and restores sensitivity to microtubule-targeting chemotherapeutic agents. Mol Cancer Ther 2009.43. Wellner U, Schubert J, Burk UC, Schmalhofer O, Zhu F, Sonntag A,Waldvogel B, Vannier C, Darling D, zur Hausen A, et al: The EMT-activatorZEB1 promotes tumorigenicity by repressing stemness-inhibitingmicroRNAs. Nat Cell Biol 2009, 11(12):1487-1495.44. Zhao Z, Rivkees SA: Rho-associated kinases play a role in endocardial celldifferentiation and migration. Dev Biol 2004, 275(1):183-191.45. Baitaluk M: System biology of gene regulation. Methods Mol Biol 2009,569:55-87.46. Hornstein E, Shomron N: Canalization of development by microRNAs. NatGenet 2006, 38(Suppl):S20-24.47. Emmrich S, Putzer BM: Checks and balances: E2F-microRNA crosstalk incancer control. Cell Cycle 2010, 9(13).Le Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 10 of 1148. Sylvestre Y, De Guire V, Querido E, Mukhopadhyay UK, Bourdeau V, Major F,Ferbeyre G, Chartrand P: An E2F/miR-20a autoregulatory feedback loop.J Biol Chem 2007, 282(4):2135-2143.49. Bracken CP, Gregory PA, Kolesnikoff N, Bert AG, Wang J, Shannon MF,Goodall GJ: A double-negative feedback loop between ZEB1-SIP1 andthe microRNA-200 family regulates epithelial-mesenchymal transition.Cancer Res 2008, 68(19):7846-7854.50. Bendoraite A, Knouf EC, Garg KS, Parkin RK, Kroh EM, O’Briant KC,Ventura AP, Godwin AK, Karlan BY, Drescher CW, et al: Regulation of miR-200 family microRNAs and ZEB transcription factors in ovarian cancer:evidence supporting a mesothelial-to-epithelial transition. Gynecol Oncol2010, 116(1):117-125.51. Wang H, Garzon R, Sun H, Ladner KJ, Singh R, Dahlman J, Cheng A,Hall BM, Qualman SJ, Chandler DS, et al: NF-kappaB-YY1-miR-29 regulatorycircuitry in skeletal myogenesis and rhabdomyosarcoma. Cancer Cell2008, 14(5):369-381.52. Chang TC, Yu D, Lee YS, Wentzel EA, Arking DE, West KM, Dang CV,Thomas-Tikhonenko A, Mendell JT: Widespread microRNA repression byMyc contributes to tumorigenesis. Nat Genet 2008, 40(1):43-50.53. Mott JL, Kurita S, Cazanave SC, Bronk SF, Werneburg NW, Fernandez-Zapico ME: Transcriptional suppression of mir-29b-1/mir-29a promoterby c-Myc, hedgehog, and NF-kappaB. J Cell Biochem 2010.54. Burk U, Schubert J, Wellner U, Schmalhofer O, Vincan E, Spaderna S,Brabletz T: A reciprocal repression between ZEB1 and members of themiR-200 family promotes EMT and invasion in cancer cells. EMBO Rep2008, 9(6):582-589.55. Liu Y, Xin Y, Ye F, Wang W, Lu Q, Kaplan HJ, Dean DC: Taz-tead1 links cell-cell contact to zeb1 expression, proliferation, and dedifferentiation inretinal pigment epithelial cells. Invest Ophthalmol Vis Sci 2010,51(7):3372-3378.56. Tan LP, Seinen E, Duns G, de Jong D, Sibon OC, Poppema S, Kroesen BJ,Kok K, van den Berg A: A high throughput experimental approach toidentify miRNA targets in human cells. Nucleic Acids Res 2009, 37(20):e137.57. Krek A, Grun D, Poy MN, Wolf R, Rosenberg L, Epstein EJ, MacMenamin P,da Piedade I, Gunsalus KC, Stoffel M, et al: Combinatorial microRNA targetpredictions. Nat Genet 2005, 37(5):495-500.58. Koscielny G, Le Texier V, Gopalakrishnan C, Kumanduri V, Riethoven JJ,Nardone F, Stanley E, Fallsehr C, Hofmann O, Kull M, et al: ASTD: TheAlternative Splicing and Transcript Diversity database. Genomics 2009,93(3):213-220.59. Zhang Y, Lv J, Liu H, Zhu J, Su J, Wu Q, Qi Y, Wang F, Li X: HHMD: thehuman histone modification database. Nucleic Acids Res 2010, , 38Database: D149-154.60. Hu M, Yu J, Taylor JM, Chinnaiyan AM, Qin ZS: On the detection andrefinement of transcription factor binding sites using ChIP-Seq data.Nucleic Acids Res 2010, 38(7):2154-2167.doi:10.1186/1471-2105-12-67Cite this article as: Le Béchec et al.: MIR@NT@N: a frameworkintegrating transcription factors, microRNAs and their targets to identifysub-network motifs in a meta-regulation network model. BMCBioinformatics 2011 12:67.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitLe Béchec et al. BMC Bioinformatics 2011, 12:67http://www.biomedcentral.com/1471-2105/12/67Page 11 of 11


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items