RIPSeeker: a statistical package for identifying protein-associated transcripts from RIP-Seq experiments Yue Li1, Dorothy Yanling Zhao 2, Jack Greenblatt2,3 and Zhaolei Zhang1,2,3 Department of Computer Science, 2Department of Molecular Genetics, 3Banting and Best Department of Medical Research, University of Toronto Molecular Cell Molecular Cell Molecular Cell Molecular Cell e Polycomb-Bound RNAs Noncoding RNA (ncRNA) biogenesis (Left; Mercer et al. 2009); long ncRNA partners with polycomb repressive complex 2 (PRC2) in gene regulation (Right; PROGRESS INSIGHT REVIEW Margueron et al. 2011) RIPSeeker workflow: Bidirectional transcript 0 ncRNA binding 5 kb Pax6 DNA binding EZH1/2 3UTR associated transcript SET JARID2 EED PTM binding SUZ12 Gapped-‐ Alignment Object Molecular Cell Automa7c bin size selec7on chr1 Y Y Y Y Y YY Y Y Y Y Markers Markers Markers C C adaptor 5’, 3’ ligation PCR adaptor Inferring RIP regions, p(z |X), using two-state HMM with negative binomial (NB) emission probability p(x |z = k) = N B(a , b ) on automatically discretized chromosome sequences: RT cDNA synthesis 5’, 3’ Illumina adaptor sequencing ligation 5’, 3’ adaptor ligation PCR PCR 11 12 11 WT Libraries RIP-seq lot library Technicalstatistics replicate 2,857,116 remaining 356,435 98,704Number 74,305of reads 14,886 58,099 Total (nonrepetitive fraction 200 5’ 3’ 5’ 3’ 5’ 3’ Biological replicate 1,913,612 231,880 87,128 63,958 12,085 47,108 WT RIP-seq 301,427 231,104 50,445 182,538 3,316,367 1,174,808 VOLUME 10 | MARCH 2009 | 157 Antisense Unannotated after Reads distinct Pilot Ezh2-/controlreplicate 1,491,715 73,691 17,424 13,86574,305 2,998 11,294 Zd`ccXeGlYc`j_\ijC`d`k\[%8cci`^_kji\j\im\[ Technical 2,857,116 356,435 reads 98,704 14,886 58,099 RNA braries RNA ncRNA filtering reads Total remaining d IgG Biological control replicate 486,315 1,913,612 4,888231,880 1,05087,128973 63,958 191 12,085 783 47,108 Rulebased 5 4 3 2 1 0 80 60 40 20 0 15000 10000 5000 0 RIPSeeker MACS QuEST HPeak Cuffdiff Rulebased 100 80 60 40 20 0 80 60 40 20 0 1500 1000 500 0 MACS QuEST HPeak Cuffdiff Rulebased protein_coding (22702) none_protein_coding (23500) lincRNA (1637) pseudogene (542) retrotransposed (263) antisense (1389) retained_intron (5213) non_coding (15) sense_intronic (81) RIPSeeker bivalent_mm8liftOver2mm9 (2704) imprint (91) lincRNA_Guttman2011 (2127) PRC2_binding_sites_mm8liftOver2mm9 (1805) lincRNA_PRC2_Guttman2011 (34) oncogenes (535) tumor_suppresors (973) RIPSeeker Detecting known PRC2-ncRNA: Xist, Kcnq1ot1, Meg3 11 22 22 Biological replicate 1,913,612 231,880 87,128 63,958 12,085 Total 0.90 maining Technical replicate Libraries CC lenges. Currently, no statistical tool is dedicated to RIP-Seq analysis. raction only) Unannotated Antisense 30 distinct 0.71 er Ezh2-/control 1,491,715 73,691 17,424 13,865 2,998 Biological replicate 60 WT RIP-seq 1.00 RNA RNA ncRNA Ezh2-/- control reads 0.27 ering gG control 486,315 4,888 1,050 973 191 V. Conclusion: Pair-wise comparison of shared peaks: Cuffdiff Rulebased HPeak RIPSeeker MACS QuEST Cuffdiff 100% 2% 4% 4% 4% 2% Rulebased 23% 100% 28% 26% 23% 12% Antisense Unannotated distinct after Pilot T RIP-seqEzh2-/- control 182,538 3,316,367 1,491,715 1,174,808 HPeak 26% 20% 100% 26% 19% 3% 73,691 301,427 17,424231,104 13,865 50,445 2,998 11,294 RNA Libraries RNA ncRNA reads filtering reads IgG control 486,315356,435 4,888 98,7041,050 74,305 973 191 783 chnical replicate 2,857,116 14,886 58,099 RIPSeeker 45% 32% 55% 100% 40% 6% II. Motivation: E F G ological replicate 1,913,612 231,880 90 87,128 63,958 12,085 47,108 WT RIP-seq 301,427 231,104 50,445 182,538 MACS 54% 41% 39% 58% 100% 12% 3,316,367 1,174,808 fraction of reads (nonrepetitive only) Libraries NumberCC LINE measures genome-wide protein-RNA interactions. Despite similarity h2-/- controlRIP-Seq1,491,715 73,691 17,424 13,865 2,998 11,294 Technical replicate 2,857,116 356,435 98,704 74,305 14,886 58,099 LTR QuEST 81% 69% 64% 71% 71% 100% E ads F G 60 G control 486,315 973 unique191 783 chalshared with ChIP- and4,888 RNA-Seq, RIP-Seq properties and WT RIP-seq 1.00 901,050 presents of all reads Cuffdiff 21 21 22 HPeak 12 12 21 QuEST Total number of peaks of reads (nonrepetitive fraction only) Unannotated Antisense distinct after Number Pilot 301,427 231,104 50,445RNA 182,538 3,316,367 Reads 1,174,808 RNA ncRNA reads filtering reads longer than 200 nucleotides that have little or X chromosomes in female mammals is inactivated. -coding capacity. Long ncRNAs can regulate X inactivation occurs so that females produce the same 3 4 6 | N AT U R E | VO L 4 6 9 | 2 0 JA N UA RY 2 0 1 1 ession through a diversity of mechanisms. dosage of gene products from the chromosome as males. ©X2011 Macmillan Publishers Limited. All rights reserved ulative % of all reads K4K36PolII Total peaks identified by comparison methods (-) ot library statistics intergenic IV. Results: analyzing PRC2 RIP-seq dataset Markers D 5% PCR Gel purification PCR GelComparison purification of RIP-Seq with ChIP-Seq and RNA-Seq: PCR h2 C cation cation α-Ezh2 h2 D T noncoding Gel purification ligation C Markers Y WT MarkersMarkers Markers Y Ezh2-/WT Markers Ezh2-/- WT Ezh2-/- C Illumina sequencing WT Markers Ezh2-/WT Markers MarkersWT Markers Ezh2-/- Ezh2-/- WT WT Bioinformatic 5’, 3’ analysis Y Bioinformatic purification analysis YY B PCR testing PCR PCR 5’, 3’ adaptor ligation Y Y B cDNA RNA extraction synthesis Y Y YY Y 5’, 3’ RT 5’, 3’ adaptor 5’, 3’ adaptor ligation adaptor NA Immuno- ligation cDNA YRT xtraction ligation precipitation Experimental synthesis RT cDNA validation; synthesis RT Functional 10% Features Ribonucleoprotein ImmunoPrecipitation (IP) followed by high-throughput Sequencing (RIP-Seq) (Zhao et al., 2010): Y coding Feature% RNAs Genome-wide Polycomb-Bound Genome-wide Polycomb-Bound RNAs RNAs Polycomb-Bound RNAs Genome-wide Polycomb-Bound Genome-wide cDNAY Y synthesis threeUTR 15% SigTest AEBP2 Several publications have reported the genome-wide localization of RbAp46/48 bin count Figure 1 | Genomic organization of coding and non-coding transcripts. of long non-coding transcripts (orange) that are associated with paired box Yes H3K27me3 in various cell lines and organisms, with some divergent Nature Reviews | Genetics posterior d ecoding f rom H MM Intronicdiagram transcript Schematic illustrating the complexity of the interleaved networks gene 6 (Pax6; purple). 5 kb results depending on the methodology used and0 the model analysed. A conservative estimation is that PRC2 targets represent at least 10% of the genes intoESrecruit cells43 . PRC2 specifically at —an and targets for Pax6 ncRNAs RNA binding proteins, resides evolution; observation supported by the antisense to introns, and they could similarly 34 Histone one ofdeposition the largest protein the mamfinding that functional repeat sequence regulate splicing H3K27me3 — theclasses Hox in genes and numerous genes encoding PTM. 44–46 malian proteome,regulators to gene promoters hugely domains are a common characteristic of Alternatively, the annealing of ncRNAbinding binding other developmental . Interestingly, in human cancer cells, -associated transcript expands the regulatory repertoire 3UTR associated transcript Mul7hits HMM posterior decoding and NB NB mixture model (G+C)-rich available several known long ncRNAs8. can target protein effector complexes to the the PRC2 component SUZ12 is mainly enriched at the promoters of genes No DNA mRNA bindingtranscript in a manner analoto the transcriptional programme29. sense exists? parameter op7miza7on ini7aliza7on 47 encodingLong glycoprotein and proteins . Further ncRNAs also act immunoglobulin-like as co-factors Post-transcriptional regulation. The abilgous to the targeting of the RNA-induced transcripts. studies of longare non-coding transcripts (orange) that areT.this associated with paired Mercer, R., Dinger, M. E., & Mattick, J. S.box (2009). Figure 3 | Thesilencing many interactions of PRC2 with chromatin. Schematic required to determine whether a Reviews consequence of to modulate transcription factor activity. ity ofisncRNAs to recognize complementary complex (RISC) to mRNAs by Nature | Genetics 2 ved networks thegene 6 (Pax6; representation of the PRC2 at chromatin. For example, in mice, thealterations ncRNA Evf2of is cancer sequences allows highly interac- siRNAs. RNA holoenzyme duplexes resulting from the Putative interactions with genetic andpurple). epigenetic cells oralso whether it isBackground aspecific either DNA orannealing histones that could explaintranscripts PRC2 recruitment are highlighted. transcribed from an ultraconserved tions that are amenable to regulating various of complementary or reflection of the cancer-cell origin. distal • export as wig, bed, etc Nuclear enhancer and recruits the binding and in H3K27me3 steps in the post-transcriptional even of long ncRNAs with extended internal Protein In Drosophila, domains enriched were found processing • (live) visualiza7on Viterbi predic7on Detect RIP action of by the the transcription factor DLX2 to of mRNAs, including their splicing, editing, hairpins can be processed into endogenous n; an observation supported antisense to introns, and they could similarly lysate to cover large regions of the genome, usually exceeding 10 kiloPRC2 recruitment Nuclear A gene beads GRanges Object Nuclear this same expression translation and degradation. siRNAs to silence expression, raising Protein α-Ezh2 48,49 enhancer to induce • (live) annota7on of enriched bins regions Protein hat functionalNuclear repeat sequence regulate splicing 34. transport, Protein RNA bases (kb) . In mammals, two Exactly how mammalian PRC2 is recruited to chromatin is not clear. 30different types of binding pattern of adjacent protein-coding genes (FIG. 2c). Most mammalian genes express antisense the possibility that many long ncRNAs feed RNA response eleysate cDNA lysate lysate Antibody Immunoare a common characteristic of Alternatively, the annealing of ncRNA A beads • other R func7ons 26 A beads A beads have been reported for PRC2 or H3K27me3: some very large domains In Drosophila, DNA sequences known as Polycomb Many similar enhancers are transcribed transcripts, which might constitute a class into RNA silencing pathways . α-Ezh2 α-Ezh2 α-Ezh2 RNA RNA RNA 8 RNA nown long ncRNAs .in cells can target protein complexes toImmunothe RNA RNA cDNA cDNA cDNA Antibody which they are active — this couldeffector of ncRNA that particularly adept at There are probably many other functions Antibody Antibody of more thanin 100 kb such as those containing the HoxisImmunoloci, andprecipitation ments (PRE) are targets for PcG protein recruitment when inserted extraction synthesis ImmunoRT incubation RNA Antibody Immuno34 41,45,47,50 3,6 Y Y sense mRNA transcript in a manner analobe a general strategy for regulating the kilobases regulating mRNA dynamics . at exogenous ofloci long ncRNAs awaiting discovery. For Y Y some smaller domains covering a few . H3K27me3 . Genetic experiments led to the identification extraction synthesis 5’, 3’ RT of RT extraction incubation synthesis RT extraction precipitation incubation synthesis 27 precipitation incubation extraction Incubation precipitation precipitation expression of key developmental genes . Antisense ncRNAs can mask key cisexample, the ncRNA NRON has been shown nscriptional regulation. The abil- to be centred gous to around the targeting of the RNA-induced i 5’, 3’ 5’,5’, enrichment seems the transcription start site of DNA-binding proteins that are required for PcG binding; however, 3’ 3’ Long ncRNAs can regulate RNA elements in mRNA by the formation of RNA to regulate the nuclear trafficking of the 41,51 by adaptor RNAs to recognize complementary silencing complex (RISC) to itself mRNAs promoters, but with a lower intensity at the start site (Fig. 2). genome-wide analysis showed that 36any one of these trans-acting factors adaptor polymerase (RNAP) II activity through duplexes, the case of the Zeb2 (also transcription factor NFAT , and the obseradaptor 34,41 as infrom i i k k adaptor s also allows Some highlyH3K27me3 specific interacsiRNAs. RNA duplexes resulting the is found at intergenic regions , and H3K27me3 only partially overlaps with PcG target genes. Instead, it is thought that other mechanisms, including by interaction called Sip1) antisense RNA, which complevation that many long ncRNAs are located in RNA-protein ligation RNA-protein ligation ligation Protein RNA-protein t are amenable to regulating various annealing of 52 complementary transcripts or 4 is enriched in subtelomeric regions and in long-terminal repeat a combination of these factors might be responsible for the recruitment with the initiation complex to influence ments the 5` splice site of an intron in the the cytoplasm suggests that they might have RNA-protein ligation complex 53 hecomplex post-transcriptional processing even of long ncRNAs with extended internal promoter choice. For example, in humans, 5` UTR of the zinc finger Hox mRNA Zeb2 undiscovered roles in cell biology. retrotransposons of PcG proteins. complex . A beads complex zh2 a ncRNAediting, transcribed from an upstream (REF. 35) .gene-expression Expression of the ncRNA prevents As, including their hairpins can be processed into endogenous cDNAin C+G, To splicing, understand how PRC2 can maintain specific In mammals, PRC2-targeted sequences are highly enriched 5’ 11111111122222211112222111111122221111…12211111 3’ region of the dihydrofolate reductase theexpression, splicing of anraising intron that contains an significance t, translation patterns, and degradation. siRNAs to silence gene the overall chromatin structure, in addition to H3K27me3 most of themMedical being classified as CpG islands, but these sequences alone synthesis (DHFR) locus forms a triplex 54 in the major internal entry site required for There is increasing interest the potential PCR 5’, 3’in adaptor 8 mmalian genes express antisense the possibility that many longribosome ncRNAs feed patterns, should be considered . This issue has generated a great deal do not indicate a consensus response element . Recently, two publicaExperimental detect RIP regions promoter of DHFR to prevent the binding efficient translation and expression of the involvement of ncRNAs in disease aetiology, 3’ 26 5’ Experimental amplification ligation Illumina ts, which might constitute a class into silencing pathways . Bioinformatic 31 of attention in the context of RNA ES-cell differentiation (Fig. 2).2e) ES. This cellssets a tions identified a mammalian PRE on the basis of the transcriptional co-factor TFIID ZEB2 protein (FIG. precedent owing to aberrant function of ncRNAs in of PcG complex recruitPCR hidden validation; Bioinformatic Illumina 9,63 Experimental A that is particularly adept Therevalidation; are probablychromatin many other functions PCR are characterized a more open and flexible organization ment in Drosophila . Both reports suggested an important role for (FIG. 2d).at This by could be a widespread mechafor ncRNAs in directing the alternative splicdifferentiation and developmental processes. analysis sequencing a a a states: Bioinformatic Illumina 34nism for controlling promoter usage as thouk=1 ing ofdiscovery. mRNA to isoforms. Indeed, aYY1, number The abilitysequencing of ncRNAs to regulate associa a a g mRNA dynamics . long ncRNAswhich awaiting ForimporFunctional analysis and a higher overall rate ofoftranscription, is thought be theofmammalian orthologue of the Drosophila PRE DNA-binding PCR validation; a a a 55 64 sands of triplex exist the in eukaryotic studies have noted the prevalenceprotein of ncRNAs ated protein-coding genes might .contribute … … Functional ense ncRNAstant can for mask key cis- 32 structures example, the ncRNA NRON has been shown pluripotency . Notably, H3K4me3 mark, often associated PHO, as previously proposed RYBP, a protein that interacts analysis sequencing a a a Automatic binning testing Gel purification chromosomes . k=2 in mRNA bywith the formation of RNA to regulate the nuclear trafficking of the active transcription, was present at most, if not all, PRC2-targeted with both YY1 and PRC1, was shown to be required for PRC1 and PRC2 Functional of read counts Gel purification Long ncRNAs can also effect testing global 39,41,43,51,56 63 36 , as in the casegenes of theinchanges Zeb2 (also transcription factor NFAT , and the obserES cells, forming the ‘bivalent domain’ . Although this recruitment . Yet genome-wide analysis in mammals did not show a Glossary by interacting with basal compo56 testing Gel65purification was of initially believed to be ES-cell , bivalent domains clear overlap MicroRNA between YY1 and PcG target genes . Moreover, PRC2 is p1) antisense pattern RNA, which complevation that many specific long Adaptive ncRNAs are located in nents the RNAP II-dependent transcripGel RIP radiation z1 z2 zi-1 zi zi+1 zN 8 4 hidden … … found differentiated somatic cells, albeit atmorphological a lowerhave freunder-represented at YY1 response . Hence, so far, there is tion machinery. ncRNAs interact with e 5` splice sitehave of anbeen intron in theinpurification thethat cytoplasm suggests that they might Evolution of new or functional Single-stranded RNAs of approximatelyelements 21–23 RIP 39,43 II machinery are typically transcribed 57 characteristics in lineages that diversify in Ab variables: IgG α-Ezh2 α-Ezh2 α-Ezh2 response to nucleotides that regulate gene expression by partial quency ; theyZeb2 were alsoundiscovered found in zebrafish but are rarely detected no strong evidence for the involvement of transcription factors in the f the zinc finger HoxRNAP mRNA roles in cell biology. environmental changes or to enable colonization of new complementary base pairing to specific mRNAs. This 58III, thereby decoupling their by RNAP WT Ezh2-/WT α-Ezh2 (-) α-Ezh2 Ab IgG Drosophila . Another histone species withecological seemingly disparate Cells recruitment of PRC2 in mammals. α-Ezh2 Expression ofinthe ncRNA prevents niches. annealing inhibits protein translation and can also facilitate RIP x1 x2 xi-1 xi xi+1 xN expression from the RNAP II-dependent ntH2Az, degradation oflong the target mRNA.RIP observed functionality that co-localizes with PRC2 is the histone variant On the other hand, ncRNAs are becoming recognized as imporng of an intron that contains an Medical significance Cells WT Ezh2-/WT (-) transcription reaction they regulate. For Epigenetic Next-gen Collect read variables: Ab 3’ IgG α-Ezh2 α-Ezh2 α-Ezh2 5’ which is usually active genes (Fig. 2). Indeed, PRC2 participants in PRC2 function. In mammals, X-chromosome ribosome entry site required forassociated increasing interest inchanges the potential Heritable in phenotype caused bytant mechanisms Transvection example, Alu elements There thatwith areistranscribed nt Ab sequences IgG α-Ezh2 α-Ezh2 α-Ezh2 sequencing 1,200 - Such outside of the genomic sequence. changes might Apparent cross-talk alleles on homologous and H2Az co-localize in involvement undifferentiated ES cells, and their recruitinactivation initiates the between expression of a 17-kb ncRNA, translation and expression of the of ncRNAs in disease aetiology, in response to heat shock bind tightly to Cells WT Ezh2-/WT (-) RIP-‐Seq Data XIST, which remain through cell divisions during, for example, cellular chromosomes, in which complementation is observed 59 II to preclude the formation of active isRNAP interdependent .owing The apparent contradiction the presence coats the X chromosome in cis. Coating with XIST RNA(-) leads to a Cells WTpromoter mutations Ezh2-/WT otein (FIG. 2e).ment This sets a precedent to aberrant function of in ncRNAs in differentiation, or nt they might persist through subsequent between in one allele and structural 1,200 33 Analysis preinitiation complexes . Alu Epigenetic changes include chromatin mutationsof in the other. Transvection can cause either gene of either H3K4me3 or H2Az withelements H3K27me3 atgenerations. the of silent marked alteration chromatin structure characterized by a progreskDpromoters NAs in directing the alternative splicdifferentiation and100 developmental processes. nt contain modular domains that can indemodifications, such as histone acetylation, or chemical activation or repression. 75 kD genes in ES cells might reflect the necessary plasticity ofto these cells, but sive heterochromatinization. The inactive X chromosome becomes RNA isoforms. Indeed, a number of The ability of ncRNAs to regulate associalterations the DNA itself, such as DNA methylation. pendently mediate polymerase binding and 500 66 1,200 - and methylated atX H3K27 100 kDcontribute also result in partial leakiness of gene silencing. That PRC2 in an XIST-dependent manner . The two long chromosome inactivation ave noted thecould prevalence of ncRNAs ated protein-coding genes might repression. In light of their abundance Long A process in which one of theby two the copiesA of the 75 kD methylation H2Az co-localize is consistent with the low levels ofncRNA DNA stem–loop structures formed repeats present 5ʹ in the XIST 1,200 and distribution in the mammalian genome, 500 Transcripts longer than 200 nucleotides that have little or X chromosomes in female mammals 60,61 67,68is inactivated. at PcG these target genes in ES cells , given evolutionarily conserved RNA interactXwith PRC2 vitro , although functional domains might have no protein-coding capacity. Long ncRNAs can regulate inactivation occursin so that females produce the samefurther regions of XIST 100thekD 62 gene expression through a diversity of mechanisms. dosage of gene products an fromXIST the X chromosome as males. been shown co-optedby into other ncRNAs during exclusivity H2Az and DNA methylation . are clearly involved because transcript in which the A repeats ry 75 kD RIP 200 100 ofkD After ES-cell differentiation, a substantial fraction bivalent can still recruit PRC2 to the XIST RNA-coated X chromo500 -domains are deleted radiation MicroRNA 43,60,61 69 ChIP-Seq RNA-Seq RIP-Seq thatorlose H3K4me3 and H2Az doblot gain DNA methylation . Notably, some . Similarly, the long ncRNA KCNQ1OT1 can mediate PRC2 75 kD Coomassie stain Western of new morphological functional Single-stranded RNAs of approximately 21–23 AbREVIEWS NATURE | GENETICS VOLUME 10 | MARCH 2009 | 157 IgG α-Ezh2 α-Ezh2 α-Ezh2 200 500 - spreading in cis, thereby maintaining the imprinted expression of the genes enriched in to both H3K27me3 anotherbymark stics in lineages that diversify in response nucleotidesand that H3K9me3, regulate gene expression partialassoci73835 231524 )''0DXZd`ccXeGlYc`j_\ijC`d`k\[%8cci`^_kji\j\im\[ 50000 70 3’ 5’ This (-) 5’ 3’ Cells WT Ezh2-/WT 5’ ntal changes or toated enable colonization of new complementary base pairing to specific mRNAs. 3’ with repression, are more abundant in human fetal3’lung fibro- KCNQ1 domain . Long ncRNA3’ could also promote PRC2 binding in ds gene DNA Coomassie stain Western blot 5’ 5’ 3’ 50 71,72 niches. annealing inhibits.protein translation and canthe alsoauthors facilitate 5’ blasts (IMR90) than in human ES cells In this same study, trans as shown for the RNA HOTAIR , the expression of which from 40000 nt degradation of the target mRNA. 200 showed that H3K27me3 domains are more extended in IMR90 cells or the HOXC locus is associated with repression of 40 kb of the HOXD c + 30000 Biorep1 transcription CD4 T cells than in ES cells, and that H3K27me3 domain expansion locus. Such mechanisms could betranscription common to a large fraction of long hanges in phenotype caused by mechanisms Transvection 50 assie stain Western blot Pilot statistics Biorep2 1,200 - library - ncRNAs73of correlates with more transcriptional silencing . Altogether, these Number . Inreads light of these results, ncRNA fraction seems to beonly) a strong candithe genomic sequence. Such changes might efficient Apparent cross-talk between alleles on homologous 200 20000 (nonrepetitive ough cell divisionsresults during, for example, that cellular in which complementation is observed indicate somaticchromosomes, cells reinforce gene silencing by increasing date for PRC2 recruitment. Reads 5’ 5’ 3’ one allele tion, or they mightthe persist through subsequent between promoter mutations in and structural RNA RNA 5’ we propose a model in which the 10000 ds DNA length of H3K27me3 domains and, for a fraction of PRC2-targeted Considering this information, massie stain Western blot Pilot library statistics 3’ s. Epigenetic changes include chromatin mutations in the other. Transvection can cause either gene Total 5’ remaining reads (nonrepetitive fraction genes, by complementary silencing pathways (H3K27me3 together with sumNumber of relatively of weak interactions or low energy steps that areonly) estab781 781 ons, such as histone acetylation, or chemical activation or repression. or DNA methylation). someReads pluripotency distinct lished by each of the PRC2 holoenzyme components would function MACS QuEST HPeak Cuffdiff Rulebased RIPSeeker Unannotated Antisense to the DNA itself,H3K9me3 such as DNA methylation. PilotNot surprisingly,after 500 - the expression of which factors, could be deleterious in differentiated together to attain the necessary energy to recruit PRC2 (Fig. 3). This X chromosome inactivation RNA Libraries RNA ncRNA reads filtering reads Total NA A process infashion which one50of of the cells, are silenced in this redundant . the two copies remaining model predicts up to four steps, not necessarily consecutive, that result AA NA NA thesis DNA nthesis ynthesis fiveUTR 20% MACS Cell Genome-wideMolecular Polycomb-Bound RNAs Disambiguate mul7hits using PCL intron 0% Cell Molecular Molecular Cell Molecular Cell Molecular Cell Genome-wide Polycomb-Bound RNAs Molecular Cell Genome-wide Genome-wide Polycomb-Bound RNAs Polycomb-Bound Molecular Cell RNAs Genome-wide localization of PRC2 andPromoter-associated H3K27me3transcript 25% Features Genome-wide Polycomb-Bound RNAs de Polycomb-Bound RNAs P R O G RRNAs ESS ide Polycomb-Bound Intronic transcript exon Feature% engaged RNA Pol II during promoter escape or elongation, rather than Antisense transcripts by regulating the initiation phase of transcription. A likely possibility is that PRC2 can repress transcription by different mechanisms, and this may be gene specific. 30% SigTest BAM/ BED/ SAM Noncoding RNA • Remove duplicate alignments • return unique hits only • flag mul7hits Comparison in biological contexts of various genomic and epigenetic features: Peak% III. Methods: probabilistic inference to disambiguate multihits and derive statistical-confidence RIP regions Features I. Introduction: genome-wide identification of long noncoding RNAs interacting with chromatin regulators Features 1 only) PRC2 47,108 11,294 783 SINE Satellite LINE Simple repeat LTR Low complexity Others SINE RIPSeeker is a self-contained software package written in R and specifically tailored to efficiently analyze RIP-Seq data with statistical rigor. RIPSeeker demonstrates its sensitivity by identifying the canonical PRC2- and CCNT1-associated (not shown) ncRNA with high statistical confidence and reasonable resolution. Additionally, RIPSeeker incorporates several existing R packages to automatically annotate RIP regions via Ensembl database, perform GO enrichments, and launch UCSC genome browser with putative RIP regions as custom tracks for visualization. Because our current knowledge of protein-associated ncRNA is largely unknown (unlike TFBS), it is difficult to evaluate the specificity of RIPSeeker predictions. However, the ability to prioritize candidate genes with rigorous statistical assessment allows RIPSeeker to generate valuable information from RIP-Seq data for formulation of subsequent (more focused) experimental and computational strategy.
© Copyright 2024 ExpyDoc
