Much of the eukaryotic genome is transcribed into non-coding RNA (ncRNA), and several studies have established that a subset of these ncRNAs form ribonucleoprotein complexes that bind and regulate chromatin (Guttman and Rinn, 2012; Meller et al., 2015; Cech and Steitz, 2014). Some of the most well studied ncRNAs are those involved in dosage compensation, which include roX1 and roX2 in Drosophila and Xist in mammals. In Drosophila, roX1 and roX2 are part of the male-specific lethal (MSL) complex that coats the single male X chromosome to acetylate histone H4K16 and increase transcription (Conrad and Akhtar, 2012). In female mammals, Xist is expressed from a single X-chromosomal locus and coats the X chromosome from which it is expressed in order to silence transcription (Augui et al., 2011). Other ncRNAs, such as HOTAIR (Rinn et al., 2007; Chu et al., 2011), HOTTIP (Wang et al., 2011), and enhancer RNAs (Sigova et al., 2015), have been shown to regulate expression of specific genes by localizing to chromatin and recruiting activating or repressing proteins. Finally, repetitive ncRNA transcripts have roles at chromosomal loci essential in maintaining genomic integrity over many cell divisions, including TERRA at telomeres (Bunting et al., 2010) and alpha-satellites near centromeres (Hall et al., 2012). Despite these well-studied examples, the genomic targets of most chromatin-associated ncRNAs are unknown, and the mechanisms by which these ncRNAs regulate the epigenetic and spatial organization of chromatin remain largely unexplored.

Genomic methods for studying the localization of specific RNA transcripts include ChIRP (Chu et al., 2011), CHART (Engreitz et al., 2013), and RAP (Simon et al., 2011). These techniques use hybridization of complementary oligonucleotides to pull down a single target RNA and then next generation sequencing or mass spectrometry to identify its DNA- or protein-binding partners (Simon et al., 2011). However, de novo discovery of chromatin-associated RNAs remains limited to computational predictions (Guttman and Rinn, 2012) or association with previously known factors (Khalil et al., 2009). Nuclear fractionation allows isolation of bulk chromatin and subsequent identification of chromatin bound RNAs via sequencing, but does not provide sequence-resolved maps of RNA binding locations along the genome (Werner and Ruthenburg, 2015). To overcome these limitations, we have developed ChAR-seq, a proximity ligation and sequencing method (Figure 1A) that both identifies chromatin-associated RNAs and maps them to genomic loci (Figure 1B).

Figure 1 with 10 supplements see all

Download asset Open asset

We developed and performed ChAR-seq using Drosophila melanogaster CME-W1-cl8+ cells, a male wing disc derived cell line with a normal karyotype and well-characterized epigenome and transcriptome (Cherbas et al., 2011; Roy et al., 2010). ChAR-seq is a chromosome conformation capture method that maps genome-wide RNA-to-DNA contacts in crosslinked nuclei (Dekker et al., 2002; Rao et al., 2014). Briefly, cells are cross-linked with formaldehyde and permeabilized, then RNA is partially fragmented and soluble RNA is washed away. The chromatin-cross-linked RNA is then ligated to an oligonucleotide duplex 'bridge' molecule and reverse transcribed. Genomic DNA is then digested and ligated onto the other end of the oligonucleotide 'bridge', creating a link between chromatin-associated RNA and proximal DNA. The ligated RNA is fully converted to cDNA during second strand synthesis. Finally, the DNA is sonicated and the chimeric molecules are purified, processed, and sequenced.

To enable the capture and analysis of RNA-to-DNA contacts, the oligonucleotide bridge (see Figure 1—figure supplement 1) was designed to have several key features: (1) the 5'-adenylated end (5'App) of the bridge ensures that it is the only 5’ end competent for ligation to the 3'-ends of ssRNA by truncated T4 Rnl2tr R55K K227Q mutant ligase (Viollet et al., 2011), which cannot adenylate 5' ends (Figure 1—figure supplement 2), (2) the sequence of the bridge does not exist in the yeast, fly, mouse or human genomes and encodes a defined polarity, (3) the end of the bridge contains a restriction site for specific ligation to digested genomic DNA, and (4) the bridge is biotinylated so that it can be captured and enriched. After the bridge is ligated to RNA in situ, the molecules are stabilized by reverse transcription using Bst3.0 polymerase, which can traverse the DNA-RNA junction. The genomic DNA is then digested using the restriction enzyme DpnII, which has a median spacing of ~200 bp between sites in the fly genome. The digested genomic DNA is then ligated to the bridge adaptor using T4 DNA ligase. Second strand synthesis is completed using limiting RNase H and DNA Polymerase I. Crosslinks are then reversed followed by DNA precipitation, sonication and purification using streptavidin beads. Libraries are constructed by repairing and dA-tailing the DNA fragments, ligating TruSeq adaptors to the ends and PCR amplifying (Figure 1A).

Upon conversion of RNA-DNA contacts to a covalent chimera, the chimeric molecules were sequenced using 152 bp single-end reads. Sequencing across the bridge junction ensures identification of the RNA and DNA portions of the chimeric molecule by reading the polarity of the bridge (Figure 1B). The RNA/cDNA (Figure 1B, red) and the genomic DNA side (Figure 1B, black) of each read were computationally split and aligned to the transcriptome and genome. After post-processing for unique alignments, repeat removal, and removal of blacklisted regions, each RNA molecule was mapped to the genomic location to which it was ligated (see Materials and methods and Figure 1—figure supplement 3), resulting in 22.2 million high-confidence unique mapping events for ~16,800 RNA transcripts. All individual RNA-to-DNA contacts for a given transcript were then combined to produce a genome-wide association map for each individual transcript (Figure 1C). To ensure that ChAR-seq signal was not due to spurious bridge-to-DNA ligation, we performed a control experiment in which we added RNase A and RNase H to lysed cells before the RNA-to-bridge ligation. This RNase-treatment reduced the number of unique bridge molecules identified by six-fold, demonstrating that the vast majority of bridge ligation events are indeed RNA-dependent (Figure 1—figure supplement 4). The ChAR-seq protocol is highly reproducible, with the number normalized RNA-to-DNA contacts observed for each RNA showing high concordance between replicates (Figure 1—figure supplement 5). To estimate the specificity of our observed RNA-to-DNA ligation events and to ensure that the contacts that we observed were not due to diffusion of RNA after fragmentation, we performed a spike-in experiment wherein we added increasing amounts of exogenous RNA to the cells after lysis but before bridge ligation. After lysis and RNA fragmentation, we recovered and quantified the total soluble RNA from the supernatant, then spiked-in purified in vitro transcribed RNA fragments (~200 nt) from commonly used protein expression vectors (MBP, HALO and GFP). The spike-in controls were added at 0.1%, 1% and 10% of the total, recoverable RNA by mass. Though we see a clear concentration-dependent increase in the number of false positives, even in the scenario where we spiked in RNA at 10% of the total soluble RNA we observed fewer than 0.5% false positive RNA-to-DNA contacts (Figure 1—figure supplement 6), which compares favorably to the false positive rates in related RNA-DNA mapping methods (Li et al., 2017; Sridhar et al., 2017).

Only the 3'-hydroxyl of each RNA is available for ligation to the bridge, thus the polarity of each RNA molecule with respect to its transcriptional direction can be determined by its orientation with respect to the bridge. The majority (85% of total) of the RNAs captured in our assay were sense, with the largest single subtype represented by sense-stranded mRNA (32% of total), owing to the capture of nascent transcripts (Figure 1—figure supplement 7). Most of the chromatin-associated antisense transcripts that we identified arose from ncRNA or intronic regions. In fact, 96% of the antisense mRNAs were intronic in origin with 64% of these originating from a single 119 kb gene (CG42339), suggesting the presence of unannotated ncRNAs in this region. The remaining chromatin-associated RNA detected in our assay arose from non-protein coding transcripts (see Figure 1—figure supplement 7), of which 18% were small nucleolar RNA (snoRNA) and 19% were small nuclear RNA (snRNA).

ChAR-seq generated RNA-to-DNA contacts can be aggregated (Figure 1C, see e.g., Total RNA), grouped by RNA class (Figure 1C, see e.g., mRNA) or viewed individually (Figure 1C). Individual RNAs mapped by ChAR-seq generally fell into one of three classes. In the first class, RNAs were found around the locus from which they are transcribed (Figure 1C, see, e.g., Hsromega, chinmo, ten-m). In the second class, RNAs were found bound to chromatin in trans, generally distributed across most or all of the genome, often in addition to a peak around the gene body from which the RNA is transcribed (Figure 1C, see e.g., snRNA:U2, snRNA:7SK). In a third class, RNAs that are part of the dosage compensation complex (Figure 1C, see roX1 and roX2) were enriched on and coat the X chromosome. To investigate this first class of RNAs, we compared aggregated RNA-to-DNA contacts with data from nascent transcription sequencing using PRO-seq (Kwak et al., 2013), and observed qualitative agreement between PRO-seq and ChAR-seq data sets (Figure 1D, see PRO-seq and mRNA). Nevertheless, many RNA-to-DNA contacts in our dataset are associated in trans to genomic regions outside of the gene body from which the RNA is transcribed (Figure 1—figure supplement 8).

To determine if the ChAR-seq protocol disrupts genome organization, we omitted the bridge to produce a mock-treated sample. We next biotinylated the DpnII-digested ends in our mock-sample and ligated them, essentially preparing a Hi-C library (‘Hi-C/Mock-ChAR’) (see Materials and methods for details) from the mock-treated sample (Figure 1E, bottom). Topologically associated domains (TADs) are preserved in the mock-treated sample and the DNA-DNA contacts show high correlation with previously published CME-W1-cl8+ cell Hi-C (Ramírez et al., 2015) (Figure 1—figure supplements 9–10). Thus, the three-dimensional genome organization is largely preserved in our protocol. Furthermore, the profile of RNA-to-DNA contacts detected by ChAR-seq was distinctly different from that of DNA-to-DNA contacts in our mock-treated Hi-C library (Figure 1E and Figure 1—figure supplement 9), indicating that ChAR-seq signal is not simply a byproduct of DNA-DNA contacts.

ChAR-seq data can also be visualized in a two-dimensional contact plot, where the genomic locus from which the RNA is transcribed is represented on the y-axis in linear genome coordinates, and the x-axis defines the genomic location where each RNA was bound. These plots provide a useful overview visualization for of the entire dataset. When we generated these contact plots for ncRNA (Figure 2A), mRNA (Figure 2B) and snRNA (Figure 2C), we observed strong horizontal lines that represent RNA transcripts that are transcribed from a single locus but are found distributed throughout the genome (class II), or in the special case of roX1 and roX2, specifically along the X chromosome (class III). Furthermore, RNAs found at sites from which they are transcribed clustered tightly along the diagonal, a feature most pronounced for mRNAs (class I) (Figure 2B). Many of the RNAs we found distributed broadly across the genome are transcription associated small nuclear RNAs (snRNAs) (Figure 2C). One of these, snRNA:7SK, is an abundant snRNA that functions as a scaffold for a transcriptional regulatory ribonucleoprotein complex that includes p-TEFb, Hexim and LARP7. Other broadly distributed snRNAs are components of the spliceosome (e.g., snRNA:U2) which largely functions co-transcriptionally (Perales and Bentley, 2009).

Figure 2 with 2 supplements see all

Download asset Open asset

To identify RNAs that are highly enriched for chromatin interactions, we plotted the normalized cumulative distribution of the number of sense contacts observed for each gene (Figure 2D). The majority of the RNAs in our dataset (14,860 out of 16,812, 88%) had fewer than 10 contacts per kilobase million reads (CPKM) (Figure 2D) and were excluded from further analysis. The remaining 1952 RNAs (12%) accounted for 83% (18.5 million) of all chromatin contacts in our data set. To estimate the contribution of total RNA abundance to this interaction signal, we performed RNA-seq for the CME-W1-cl8+ cell line and compared RNA expression levels with RNA-to-DNA contacts identified by ChAR-seq (Figure 2E, Supplementary file 1). We observed a correlation between RNA expression level and chromatin-RNA contacts; however, a cluster of RNAs clearly generated more chromatin interactions that would be expected from the overall expression levels (Figure 2E). Using both the length and read normalized contacts (CPKM) and the fold-enrichment over RNA expression as measured by RNA-seq, we identified 138 RNAs that had more than 100 CPKM and were enriched more than ten-fold, though many were enriched by 2–5 orders of magnitude (Figure 2E, red symbols; Figure 2—figure supplement 1). Notably, we observe good concordance between the RNAs identified using this methodology between replicates (Figure 2—figure supplement 2).

We developed ChAR-seq using the male WME-cl8+ line, reasoning that the ncRNAs roX1 and roX2 would serve as an internal positive control. Both roX1 and roX2 are part of the MSL2 complex, which binds across the X-chromosome in male flies to recruit chromatin-modifiers that increase transcriptional output (Figure 3A) (Lucchesi and Kuroda, 2015). Indeed, ChAR-seq data showed roX1 and roX2 to be 7.6-fold (p-value<10⁻¹⁰) and 8.1-fold (p-value<10⁻¹⁰) enriched for interactions on the X chromosome, respectively (Figure 3B,C, Figure 3—figure supplement 1). In contrast, female flies express Sex lethal (Sxl), which binds to msl2 mRNA to prevent its translation, blocking assembly of the MSL2 complex (Lucchesi and Kuroda, 2015). Importantly, roX1 and roX2 require MSL2 for X-chromosome specific localization (Lucchesi and Kuroda, 2015), therefore female cells should lack detectable spreading of these ncRNAs along the X-chromosome. When we performed ChAR-seq in a female Drosophila melanogaster cell line, Kc167, we did not detect any significant roX2 localization on the X chromosome (Figure 3D) but observed excellent agreement in interaction signal from other RNAs across both cell lines (Figure 3—figure supplement 2, Figure 3C male, CME-W1-cl8+ and Figure 3D, female, Kc167, see e.g., snRNA:7SK and Hsromega).

Figure 3 with 2 supplements see all

Download asset Open asset

High-resolution maps of roX1 and roX2 localization have previously been generated using ChIRP-Seq, which hybridizes probes against a known RNA and pulls down the associated chromatin for sequencing (Chu et al., 2011; Quinn et al., 2014). Comparing ChIRP-seq to ChAR-seq for both roX1 and roX2 (Figure 3E), we found that DNA contact locations were in surprisingly good agreement despite the fact that ChAR-seq reads are spread across all RNAs while ChIRP-seq reads map the specific RNA target, resulting in a large disparity in the effective sequencing depth between the methods. In ChIRP-seq, virtually all of the signal is attributable to interactions between chromatin and the target RNA. In contrast, ChAR-seq captures all RNA and DNA contacts, so that any given target RNA will comprise a subset of the total RNA-chromatin contacts in the dataset. In the case of roX1 and roX2, we observed 32,308 and 87,453 contacts, representing 0.1% and 0.36% of the ChAR-seq dataset. In contrast, the ChIRP-seq datasets plotted in Figure 3E represent ~24M and ~21M reads for roX1 and roX2, respectively. This indicates that ChAR-seq can identify RNA peaks along chromatin with high sensitivity for a given RNA.

To more quantitatively compare ChAR-seq to ChIRP-seq, we compared the signal-to-noise ratio (SNR) for roX1 and roX2 binding to DNA in each assay (Figure 3F). We defined signal as the most densely bound regions on chrX, whereas we defined noise as binding events distributed randomly on autosomes (see Materials and methods for details). We found that ChAR-seq roX1 SNR was much higher than that of ChIRP-seq (49.8 vs 3.9), and ChAR-seq roX2 SNR was about two-fold higher than that of ChIRP-seq (45.2 vs 22.0), despite having 2 orders of magnitude fewer reads in ChAR-seq. Altogether, these data suggest that ChAR-seq has excellent sensitivity and sufficient signal-to-noise to characterize accurate chromatin-binding events for individual RNAs.

The resolution with which we can measure the localization of an RNA to a given genomic site constrains our ability to assess its potential modes of action. To measure the accuracy of ChAR-seq measurements of RNA interaction with DNA, we compared ChAR-Seq data to ChIRP-seq data to estimate the base-pair resolution of ChAR-Seq. We expected this resolution to be bounded ––in part––by the local DpnII cut frequency and the number of contacts for any given RNA. We divided the X chromosome into evenly sized bins and calculated correlation coefficients between ChIRP-seq and ChAR-seq datasets at increasing bin sizes for both roX1 and roX2 (Figure 3G). Using this method, we noted a bi-phasic increase of the correlation coefficient, corresponding to a minor plateau around 200 bp and a major plateau at ~25 kbp. The minor plateau is likely due to the DpnII distribution bias in the ChAR-seq tracks, while the major plateau is an estimate of the resolution of our assay, which is on the order of other proximity-ligation sequencing assays like Hi-C (van Berkum et al., 2010).

To test if we could identify the functional roles for our most highly enriched RNAs, we clustered the snRNA class of RNAs based on their genomic contacts. These snRNAs collectively comprised 23% of all the RNA-to-DNA contacts in our dataset (Figure 4A) and are a substantial component of the spliceosome, a multi-megadalton ribonucleoprotein complex that catalyzes pre-mRNA splicing (Zhou et al., 2002; Will and Lührmann, 2011). The composition and conformation of the spliceosome is highly dynamic, though two dominant species exist in eukaryotes: the major spliceosome comprised of U1, U2, U4, U5 and U6 snRNAs, and the minor spliceosome comprised of U4:atac, U6:atac, U5, U11, and U12 (Will and Lührmann, 2011). Many members of this class of snRNAs have highly similar gene duplication variants in the Drosophila genome. We therefore first calculated the base sequence similarity of these variants to one another and aggregated signals that were tightly clustered (Figure 4—figure supplement 1). When we then correlated genome-wide binding signal within this class, we found that the distribution patterns of the major spliceosome snRNAs U1, U2, U4, U5, U6 clustered together along with snRNA:7SK (Figure 4B), which is part of the p-TEFb complex that relieves pausing of RNA Polymerase II at promoters (Kwak and Lis, 2013) and may participate in the release of paused polymerase during RNA splicing (Barboric et al., 2009). The components of the minor spliceosome did not cluster together, likely due to their low abundance (Will and Lührmann, 2011) and consequently low representation in our dataset.

Figure 4 with 2 supplements see all

Download asset Open asset

We next reasoned that spliceosome RNAs––as part of the co-transcriptional RNA processing machinery––should also be enriched in regions of active transcription. We therefore aggregated spliceosomal RNA signals over gene bodies (Figure 4C, red lines), putative enhancers (Kvon et al., 2014) (Figure 4C, blue dashed lines) and a random distribution of genomic bins of similar size (Figure 4C, black lines). We observed an enrichment of snRNAs (7SK, U2 and U6), but not roX1 or roX2, over gene bodies (Figure 4C) with a broad peak around transcription start sites, in good agreement with ChIRP data for 7SK in mice (Flynn et al., 2016). Because active transcription is also correlated with topological boundaries in flies (Hou et al., 2012; Ulianov et al., 2016), we examined the relationship between RNA contacts and genome organization. To test whether RNA-DNA contacts are enriched at topological boundaries, we measured the DNA contact frequency for snRNA:7SK and snRNA:U2 in 20 kb windows spanning TAD boundaries and TAD centers. Both 7SK and U2 RNAs were modestly, but significantly (Wilcoxon rank-sum test) enriched at TAD boundaries (Figure 4—figure supplement 2A (top)). We also aggregated 7SK and U2 contact signals across all TAD boundaries (Figure 4—figure supplement 2A (bottom), red lines) and a random distribution of identically sized genomic bins (Figure 4—figure supplement 2A (bottom), black lines), and found that these RNAs were ~1.4 fold enriched over boundaries. We also noted that roX2 was ~2.2 fold enriched over TAD boundaries on the X chromosome (Figure 4—figure supplement 2B), consistent with previous data showing that MSL2 dosage compensation complex High Affinity Sites (HAS) are preferentially found at TAD boundaries (Ramírez et al., 2015). Examination of chromatin accessibility (ATAC-seq) over TAD boundaries (Figure 4—figure supplement 2B) showed that open chromatin is enriched at TAD boundaries, supporting the idea that TAD boundaries are transcriptionally active in flies.

In contrast to the small number of well-defined and well-characterized snRNAs involved in splicing, there are more than 200 snoRNAs in flies (Huang et al., 2005) that are significantly divergent in sequence and, surprisingly, were highly represented in our dataset (Figure 1—figure supplement 7 and Figure 2—figure supplements 1–2). Most of these snoRNAs have either unknown function or are computationally identified and indirectly implicated in the maturation and modification of ribosomal rRNA (Huang et al., 2005).

To determine if our enriched chromatin-associated RNAs, in particular snRNAs and snoRNAs, might localize to euchromatic or heterochromatic states or with specific transcription factors, we cross-correlated our ChAR-seq signal against modENCODE datasets available for the CME-W1-cl8+ cell line. To normalize the signals for comparison, we first calculated the expected contacts per 2 kb bin for each RNA under a uniform distribution, based on the total number of genome-wide contacts for each RNA and the number of DpnII sites per bin. This null model was then used to calculate the log2 ratio of the observed to the expected contacts per bin for each RNA, which was then transformed into a z-score ((x-μ)/σ) based on the whole-genome mean (μ) and standard deviation (σ). Similarly, we re-binned the modENCODE tracks, removed bins that did not contain a DpnII site, and transformed the log2 mean-shift values to a z-score. We then calculated the pair-wise Pearson correlation coefficients between each signal track, and then clustered the data (Figure 4D). We observed discrete clustering of roX1 and roX2 with known dosage compensation complex factors, MOF, the histone modifications H4K16ac and H3K36me3 (Bell et al., 2008), and JIL-1 kinase (Bai et al., 2004), validating this analytical approach (Figure 4D). Beyond this sub-cluster of dosage compensation factors, the remainder of the chromatin-associated RNAs fell into two distinct and anti-correlated categories: those associated with active chromatin and transcription (e.g., RNAPII, H4K8ac, H3K18ac) or heterochromatin (e.g., HP2, H3K9me3, HP1a) (Figure 4D). In particular, we note that snRNA:U2 and snRNA:7SK cluster tightly with the transcription-associated chromatin marks, while many of the snoRNAs and minor spliceosome snRNAs that we identified are associated with heterochromatin, likely due to co-localization of heterochromatin factors to the nucleolus. Interestingly, snRNA:U5, a component of both the major and minor spliceosome, has variants that clearly cluster with either transcriptionally active chromatin (63BC) or heterochromatin (23D, 38ABa, and 34A). Previous work has shown that the snRNA:U5:38Aba variant (Figure 4D, heterochromatin cluster) exhibits a unique tissue-specific expression profile with the greatest abundance in neural tissue, which led the authors to propose isoform-dependent functions in alternative splicing (Chen et al., 2005). The differential clustering that we observe for snRNA:U5, and between major and minor spliceosome snRNAs, for euchromatin and heterochromatin might reflect such isoform-specific functions of the spliceosome in different chromatin states.

ChAR-seq maps the chromosomal binding sites of all chromatin-associated RNAs, independent of whether they are associated as nascent transcripts or bound as part of ribonucleoprotein complexes (RNPs). In this way, ChAR-seq can be thought of as a massively parallelized de novo RNA mapping assay capable of generating hundreds to thousands of RNA-binding maps. ChAR-seq also detects multiple classes of chromatin-associated RNAs. ChAR-seq preserves the three-dimensional structure of the genome and provides an RNA-DNA interaction matrix that complements DNA-DNA proximity measurements using Hi-C. We validated ChAR-seq using chromosome-specific ncRNAs roX1 and roX2 associated with dosage compensation. The comparison between ChAR-seq and ChIRP-seq, which vary dramatically in the sequencing depth needed to analyze a specific RNA, highlights the utility of ChAR-seq as a de novo chromatin-associated RNA discovery tool. ChAR-seq also maps nascent RNAs found at the loci from which they are transcribed.

ChAR-seq is similar to a recently published method (Sridhar et al., 2017), but has two key distinctions. First, proximity ligations are performed in situ in intact nuclei, which reduces nonspecific interactions (Rao et al., 2014). Second, ChAR-seq uses long single-end reads to sequence across the entire junction of the ‘bridge’, ensuring that RNA-to-DNA contacts are mapped with high confidence and reporting on the polarity of the bridge-ligated RNA.

While this work was in revision, a genome-wide RNA-DNA contact method—GRID-seq—was published that also uses proximity ligation of a directional bridge to RNA and DNA (Li et al., 2017). One key difference is that GRID-seq uses a restriction enzyme to cut cDNA and gDNA fragments 19–23 bp distal to the bridge, which allows size selection of molecules that contain both RNA and DNA. This likely reduces the number of uninformative molecules sequenced. A disadvantage, however, is that the resulting reads are 19–23 bp, which have a greater chance of falsely mapping during genome alignment than the 20–100 bp RNA and DNA fragments obtained with ChAR-seq. Although their details differ, ChAR-seq and GRID-seq appear to have similar ability to detect roX2 binding to chrX in Drosophila (see Figure 3—figure supplement 1B).

We used ChAR-seq to discover and map several dozen ncRNAs that are pervasively bound across the genome. Many of these ncRNAs are components of ribonucleoprotein complexes associated with transcription elongation (snRNA:7SK), splicing (snRNA:U2, etc) and RNA processing (snRNAs, snoRNAs and scaRNAs). Interestingly, more than half of the chromatin-associated RNAs identified based on our enrichment criteria are snoRNAs, most of which––but not all––correlate with heterochromatin. Generally, snoRNA ribonucleoproteins (snoRNPs) use intermolecular base pairing to direct chemical modification of the 2'-hydroxyl groups or the isomerization of uridines to pseudouridine (Cech and Steitz, 2014) and snoRNAs are known abundant components of chromatin in both flies (Schubert et al., 2012) and in mice (Meng et al., 2016). Despite their abundance and the their known role in RNA modification, we do not yet understand the functions of these modifications, or the implication of snoRNAs and scaRNAs chromatin association in cells (Cech and Steitz, 2014). We also demonstrate that ChAR-seq can be used with orthogonal genome-wide datasets to identify and classify RNAs that are associated with specific chromatin states (e.g., euchromatin vs heterochromatin). We expect this approach will be particularly useful in organisms that use lncRNAs such as HOTAIR, HOTTIP and BRAVEHEART as scaffolds for ribonucleoproteins that regulate facultative heterochromatin. Finally, we observed enrichment of transcription-associated RNAs with TAD boundaries, reflecting a potential role for active transcription in the topological organization of the genome. This observation is consistent with previous observations that active transcription is a stronger predictor for TAD partitioning in flies (Hou et al., 2012; Ulianov et al., 2016) than CTCF and cohesin, the prototypical TAD boundary markers in mice and humans (Merkenschlager and Nora, 2016).

We anticipate that ChAR-seq will be a powerful new high throughput discovery platform capable of simultaneously identifying new chromatin-associated RNAs and mapping their chromatin binding sites (and associated epigenetic chromatin states), all of which will be particularly useful in comparing ‘epigenomic’ changes that coincide with cellular differentiation or tumorigenesis.

1. 1. Bell JC
   2. Jukam D
   3. Teran NA
   4. Risca VI
   5. Smith OK
   6. Johnson WL
   7. Skotheim J
   8. Greenleaf WJ
   9. Straight AF

   (2017) Chromatin-associated RNA sequencing (ChAR-seq) maps genome-wide RNA-to-DNA contacts

   Publicly available at the NCBI Gene Expression Omnibus (accession no: GSE97131).

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE97131

1. 1. Quinn JJ
   2. Qu K
   3. Chang HY

   (2014) Domain ChIRP reveals the modularity of long noncoding RNA architecture, function, and target genes

   Publicly available at the NCBI Gene Expression Omnibus (accession no: GSE53020).

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE53020
2. 1. Ramírez F
   2. Lingg T
   3. Toscano S
   4. Lam KC
   5. Georgiev P
   6. Chung HR
   7. Lajoie BR
   8. de Wit E
   9. Zhan Y
   10. de Laat W
   11. Dekker J
   12. Manke T
   13. Akhtar A

   (2015) High-Affinity Sites Form an Interaction Network to Facilitate Spreading of the MSL Complex across the X Chromosome in Drosophila

   Publicly available at the NCBI Gene Expression Omnibus (accession no: GSE58821).

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE58821
3. 1. Li X
   2. Zhou B
   3. Chen L
   4. Gao L-T
   5. Li H
   6. Fu X-D

   (2017) GRID-seq reveals the global RNA-chromatin interactome

   Publicly available at the NCBI Gene Expression Omnibus (accession no: GSE82312).

   https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE82312

1. 1. Augui S
   2. Nora EP
   3. Heard E

   (2011) Regulation of X-chromosome inactivation by the X-inactivation centre

   Nature Reviews Genetics 12:429–442.

   https://doi.org/10.1038/nrg2987

   • PubMed
   • Google Scholar
2. 1. Bai X
   2. Alekseyenko AA
   3. Kuroda MI

   (2004) Sequence-specific targeting of MSL complex regulates transcription of the roX RNA genes

   The EMBO Journal 23:2853–2861.

   https://doi.org/10.1038/sj.emboj.7600299

   • PubMed
   • Google Scholar
3. 1. Barboric M
   2. Lenasi T
   3. Chen H
   4. Johansen EB
   5. Guo S
   6. Peterlin BM

   (2009) 7SK snRNP/P-TEFb couples transcription elongation with alternative splicing and is essential for vertebrate development

   PNAS 106:7798–7803.

   https://doi.org/10.1073/pnas.0903188106

   • PubMed
   • Google Scholar
4. Software

   1. Bell J

   (2017)

   Pipeline to process single end ChARseq reads to a combined, filtered and annotated matrix of RNA to DNA contacts, version 662e4bb

   Github.
5. 1. Bell O
   2. Conrad T
   3. Kind J
   4. Wirbelauer C
   5. Akhtar A
   6. Schübeler D

   (2008) Transcription-coupled methylation of histone H3 at lysine 36 regulates dosage compensation by enhancing recruitment of the MSL complex in Drosophila melanogaster

   Molecular and Cellular Biology 28:3401–3409.

   https://doi.org/10.1128/MCB.00006-08

   • PubMed
   • Google Scholar
6. 1. Bolger AM
   2. Lohse M
   3. Usadel B

   (2014) Trimmomatic: a flexible trimmer for illumina sequence data

   Bioinformatics 30:2114–2120.

   https://doi.org/10.1093/bioinformatics/btu170

   • PubMed
   • Google Scholar
7. 1. Buenrostro JD
   2. Giresi PG
   3. Zaba LC
   4. Chang HY
   5. Greenleaf WJ

   (2013) Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position

   Nature Methods 10:1213–1218.

   https://doi.org/10.1038/nmeth.2688

   • PubMed
   • Google Scholar
8. 1. Bunting SF
   2. Callén E
   3. Wong N
   4. Chen HT
   5. Polato F
   6. Gunn A
   7. Bothmer A
   8. Feldhahn N
   9. Fernandez-Capetillo O
   10. Cao L
   11. Xu X
   12. Deng CX
   13. Finkel T
   14. Nussenzweig M
   15. Stark JM
   16. Nussenzweig A

   (2010) 53BP1 inhibits homologous recombination in Brca1-deficient cells by blocking resection of DNA breaks

   Cell 141:243–254.

   https://doi.org/10.1016/j.cell.2010.03.012

   • PubMed
   • Google Scholar
9. 1. Cech TR
   2. Steitz JA

   (2014) The noncoding RNA revolution-trashing old rules to forge new ones

   Cell 157:77–94.

   https://doi.org/10.1016/j.cell.2014.03.008

   • PubMed
   • Google Scholar
10. 1. Chen L
    2. Lullo DJ
    3. Ma E
    4. Celniker SE
    5. Rio DC
    6. Doudna JA

    (2005) Identification and analysis of U5 snRNA variants in Drosophila

    RNA 11:1473–1477.

    https://doi.org/10.1261/rna.2141505

    • PubMed
    • Google Scholar
11. 1. Cherbas L
    2. Willingham A
    3. Zhang D
    4. Yang L
    5. Zou Y
    6. Eads BD
    7. Carlson JW
    8. Landolin JM
    9. Kapranov P
    10. Dumais J
    11. Samsonova A
    12. Choi JH
    13. Roberts J
    14. Davis CA
    15. Tang H
    16. van Baren MJ
    17. Ghosh S
    18. Dobin A
    19. Bell K
    20. Lin W
    21. Langton L
    22. Duff MO
    23. Tenney AE
    24. Zaleski C
    25. Brent MR
    26. Hoskins RA
    27. Kaufman TC
    28. Andrews J
    29. Graveley BR
    30. Perrimon N
    31. Celniker SE
    32. Gingeras TR
    33. Cherbas P

    (2011) The transcriptional diversity of 25 Drosophila cell lines

    Genome Research 21:301–314.

    https://doi.org/10.1101/gr.112961.110

    • PubMed
    • Google Scholar
12. 1. Chu C
    2. Qu K
    3. Zhong FL
    4. Artandi SE
    5. Chang HY

    (2011) Genomic maps of long noncoding RNA occupancy reveal principles of RNA-chromatin interactions

    Molecular Cell 44:667–678.

    https://doi.org/10.1016/j.molcel.2011.08.027

    • PubMed
    • Google Scholar
13. 1. Conrad T
    2. Akhtar A

    (2012) Dosage compensation in Drosophila melanogaster: epigenetic fine-tuning of chromosome-wide transcription

    Nature Reviews Genetics 13:123–134.

    https://doi.org/10.1038/nrg3124

    • PubMed
    • Google Scholar
14. 1. Dekker J
    2. Rippe K
    3. Dekker M
    4. Kleckner N

    (2002) Capturing chromosome conformation

    Science 295:1306–1311.

    https://doi.org/10.1126/science.1067799

    • PubMed
    • Google Scholar
15. 1. ENCODE Project Consortium
    2. The ENCODE Project Consortium

    (2012) An integrated encyclopedia of DNA elements in the human genome

    Nature 489:57–74.

    https://doi.org/10.1038/nature11247

    • PubMed
    • Google Scholar
16. 1. Engreitz JM
    2. Pandya-Jones A
    3. McDonel P
    4. Shishkin A
    5. Sirokman K
    6. Surka C
    7. Kadri S
    8. Xing J
    9. Goren A
    10. Lander ES
    11. Plath K
    12. Guttman M

    (2013) The Xist lncRNA exploits three-dimensional genome architecture to spread across the X chromosome

    Science 341:1237973.

    https://doi.org/10.1126/science.1237973

    • PubMed
    • Google Scholar
17. 1. Flynn RA
    2. Do BT
    3. Rubin AJ
    4. Calo E
    5. Lee B
    6. Kuchelmeister H
    7. Rale M
    8. Chu C
    9. Kool ET
    10. Wysocka J
    11. Khavari PA
    12. Chang HY

    (2016) 7SK-BAF axis controls pervasive transcription at enhancers

    Nature Structural & Molecular Biology 23:231–238.

    https://doi.org/10.1038/nsmb.3176

    • PubMed
    • Google Scholar
18. 1. Guttman M
    2. Rinn JL

    (2012) Modular regulatory principles of large non-coding RNAs

    Nature 482:339–346.

    https://doi.org/10.1038/nature10887

    • PubMed
    • Google Scholar
19. 1. Hall LE
    2. Mitchell SE
    3. O'Neill RJ

    (2012) Pericentric and centromeric transcription: a perfect balance required

    Chromosome Research 20:535–546.

    https://doi.org/10.1007/s10577-012-9297-9

    • PubMed
    • Google Scholar
20. 1. Heinz S
    2. Benner C
    3. Spann N
    4. Bertolino E
    5. Lin YC
    6. Laslo P
    7. Cheng JX
    8. Murre C
    9. Singh H
    10. Glass CK

    (2010) Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities

    Molecular Cell 38:576–589.

    https://doi.org/10.1016/j.molcel.2010.05.004

    • PubMed
    • Google Scholar
21. 1. Hou C
    2. Li L
    3. Qin ZS
    4. Corces VG

    (2012) Gene density, transcription, and insulators contribute to the partition of the Drosophila genome into physical domains

    Molecular Cell 48:471–484.

    https://doi.org/10.1016/j.molcel.2012.08.031

    • PubMed
    • Google Scholar
22. 1. Huang ZP
    2. Zhou H
    3. He HL
    4. Chen CL
    5. Liang D
    6. Qu LH

    (2005) Genome-wide analyses of two families of snoRNA genes from Drosophila melanogaster, demonstrating the extensive utilization of introns for coding of snoRNAs

    RNA 11:1303–1316.

    https://doi.org/10.1261/rna.2380905

    • PubMed
    • Google Scholar
23. 1. Kearse M
    2. Moir R
    3. Wilson A
    4. Stones-Havas S
    5. Cheung M
    6. Sturrock S
    7. Buxton S
    8. Cooper A
    9. Markowitz S
    10. Duran C
    11. Thierer T
    12. Ashton B
    13. Meintjes P
    14. Drummond A

    (2012) Geneious basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data

    Bioinformatics 28:1647–1649.

    https://doi.org/10.1093/bioinformatics/bts199

    • PubMed
    • Google Scholar
24. 1. Khalil AM
    2. Guttman M
    3. Huarte M
    4. Garber M
    5. Raj A
    6. Rivea Morales D
    7. Thomas K
    8. Presser A
    9. Bernstein BE
    10. van Oudenaarden A
    11. Regev A
    12. Lander ES
    13. Rinn JL

    (2009) Many human large intergenic noncoding RNAs associate with chromatin-modifying complexes and affect gene expression

    PNAS 106:11667–11672.

    https://doi.org/10.1073/pnas.0904715106

    • PubMed
    • Google Scholar
25. 1. Krzywinski M
    2. Schein J
    3. Birol I
    4. Connors J
    5. Gascoyne R
    6. Horsman D
    7. Jones SJ
    8. Marra MA

    (2009) Circos: an information aesthetic for comparative genomics

    Genome Research 19:1639–1645.

    https://doi.org/10.1101/gr.092759.109

    • PubMed
    • Google Scholar
26. 1. Kvon EZ
    2. Kazmar T
    3. Stampfel G
    4. Yáñez-Cuna JO
    5. Pagani M
    6. Schernhuber K
    7. Dickson BJ
    8. Stark A

    (2014) Genome-scale functional characterization of Drosophila developmental enhancers in vivo

    Nature 512:91–95.

    https://doi.org/10.1038/nature13395

    • PubMed
    • Google Scholar
27. 1. Kwak H
    2. Fuda NJ
    3. Core LJ
    4. Lis JT

    (2013) Precise maps of RNA polymerase reveal how promoters direct initiation and pausing

    Science 339:950–953.

    https://doi.org/10.1126/science.1229386

    • PubMed
    • Google Scholar
28. 1. Kwak H
    2. Lis JT

    (2013) Control of transcriptional elongation

    Annual Review of Genetics 47:483–508.

    https://doi.org/10.1146/annurev-genet-110711-155440

    • PubMed
    • Google Scholar
29. 1. Langmead B
    2. Salzberg SL

    (2012) Fast gapped-read alignment with bowtie 2

    Nature Methods 9:357–359.

    https://doi.org/10.1038/nmeth.1923

    • PubMed
    • Google Scholar
30. 1. Li H
    2. Handsaker B
    3. Wysoker A
    4. Fennell T
    5. Ruan J
    6. Homer N
    7. Marth G
    8. Abecasis G
    9. Durbin R
    10. 1000 Genome Project Data Processing Subgroup

    (2009) The sequence alignment/Map format and SAMtools

    Bioinformatics 25:2078–2079.

    https://doi.org/10.1093/bioinformatics/btp352

    • PubMed
    • Google Scholar
31. 1. Li X
    2. Zhou B
    3. Chen L
    4. Gou LT
    5. Li H
    6. Fu XD

    (2017) GRID-seq reveals the global RNA-chromatin interactome

    Nature Biotechnology 35:940–950.

    https://doi.org/10.1038/nbt.3968

    • PubMed
    • Google Scholar
32. 1. Lucchesi JC
    2. Kuroda MI

    (2015) Dosage compensation in Drosophila

    Cold Spring Harbor Perspectives in Biology 7:a019398.

    https://doi.org/10.1101/cshperspect.a019398

    • PubMed
    • Google Scholar
33. 1. Meller VH
    2. Joshi SS
    3. Deshpande N

    (2015) Modulation of chromatin by noncoding RNA

    Annual Review of Genetics 49:673–695.

    https://doi.org/10.1146/annurev-genet-112414-055205

    • PubMed
    • Google Scholar
34. 1. Meng Y
    2. Yi X
    3. Li X
    4. Hu C
    5. Wang J
    6. Bai L
    7. Czajkowsky DM
    8. Shao Z

    (2016) The non-coding RNA composition of the mitotic chromosome by 5'-tag sequencing

    Nucleic Acids Research 44:4934–4946.

    https://doi.org/10.1093/nar/gkw195

    • PubMed
    • Google Scholar
35. 1. Merkenschlager M
    2. Nora EP

    (2016) CTCF and Cohesin in genome folding and transcriptional gene regulation

    Annual Review of Genomics and Human Genetics 17:17–43.

    https://doi.org/10.1146/annurev-genom-083115-022339

    • PubMed
    • Google Scholar
36. 1. Perales R
    2. Bentley D

    (2009) "Cotranscriptionality": the transcription elongation complex as a nexus for nuclear transactions

    Molecular Cell 36:178–191.

    https://doi.org/10.1016/j.molcel.2009.09.018

    • PubMed
    • Google Scholar
37. Conference

    1. Petersen KR
    2. Streett DA
    3. Gerritsen AT
    4. Hunter SS
    5. Settles ML

    (2015)

    Super deduper, fast PCR duplicate detection in fastq files

    BCB '15 ACM International Conference on Bioinformatics, Computational Biology and Biomedicine. pp. 491–492.

    • Google Scholar
38. 1. Quinlan AR
    2. Hall IM

    (2010) BEDTools: a flexible suite of utilities for comparing genomic features

    Bioinformatics 26:841–842.

    https://doi.org/10.1093/bioinformatics/btq033

    • PubMed
    • Google Scholar
39. 1. Quinn JJ
    2. Ilik IA
    3. Qu K
    4. Georgiev P
    5. Chu C
    6. Akhtar A
    7. Chang HY

    (2014) Revealing long noncoding RNA architecture and functions using domain-specific chromatin isolation by RNA purification

    Nature Biotechnology 32:933–940.

    https://doi.org/10.1038/nbt.2943

    • PubMed
    • Google Scholar
40. 1. Ramírez F
    2. Lingg T
    3. Toscano S
    4. Lam KC
    5. Georgiev P
    6. Chung HR
    7. Lajoie BR
    8. de Wit E
    9. Zhan Y
    10. de Laat W
    11. Dekker J
    12. Manke T
    13. Akhtar A

    (2015) High-affinity sites form an interaction network to facilitate spreading of the MSL Complex across the X Chromosome in Drosophila

    Molecular Cell 60:146–162.

    https://doi.org/10.1016/j.molcel.2015.08.024

    • PubMed
    • Google Scholar
41. 1. Ramírez F
    2. Ryan DP
    3. Grüning B
    4. Bhardwaj V
    5. Kilpert F
    6. Richter AS
    7. Heyne S
    8. Dündar F
    9. Manke T

    (2016) deepTools2: a next generation web server for deep-sequencing data analysis

    Nucleic Acids Research 44:W160–W165.

    https://doi.org/10.1093/nar/gkw257

    • PubMed
    • Google Scholar
42. 1. Rao SS
    2. Huntley MH
    3. Durand NC
    4. Stamenova EK
    5. Bochkov ID
    6. Robinson JT
    7. Sanborn AL
    8. Machol I
    9. Omer AD
    10. Lander ES
    11. Aiden EL

    (2014) A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping

    Cell 159:1665–1680.

    https://doi.org/10.1016/j.cell.2014.11.021

    • PubMed
    • Google Scholar
43. 1. Rinn JL
    2. Kertesz M
    3. Wang JK
    4. Squazzo SL
    5. Xu X
    6. Brugmann SA
    7. Goodnough LH
    8. Helms JA
    9. Farnham PJ
    10. Segal E
    11. Chang HY

    (2007) Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs

    Cell 129:1311–1323.

    https://doi.org/10.1016/j.cell.2007.05.022

    • PubMed
    • Google Scholar
44. 1. Roy S
    2. Ernst J
    3. Kharchenko PV
    4. Kheradpour P
    5. Negre N
    6. Eaton ML
    7. Landolin JM
    8. Bristow CA
    9. Ma L
    10. Lin MF
    11. Washietl S
    12. Arshinoff BI
    13. Ay F
    14. Meyer PE
    15. Robine N
    16. Washington NL
    17. Di Stefano L
    18. Berezikov E
    19. Brown CD
    20. Candeias R
    21. Carlson JW
    22. Carr A
    23. Jungreis I
    24. Marbach D
    25. Sealfon R
    26. Tolstorukov MY
    27. Will S
    28. Alekseyenko AA
    29. Artieri C
    30. Booth BW
    31. Brooks AN
    32. Dai Q
    33. Davis CA
    34. Duff MO
    35. Feng X
    36. Gorchakov AA
    37. Gu T
    38. Henikoff JG
    39. Kapranov P
    40. Li R
    41. MacAlpine HK
    42. Malone J
    43. Minoda A
    44. Nordman J
    45. Okamura K
    46. Perry M
    47. Powell SK
    48. Riddle NC
    49. Sakai A
    50. Samsonova A
    51. Sandler JE
    52. Schwartz YB
    53. Sher N
    54. Spokony R
    55. Sturgill D
    56. van Baren M
    57. Wan KH
    58. Yang L
    59. Yu C
    60. Feingold E
    61. Good P
    62. Guyer M
    63. Lowdon R
    64. Ahmad K
    65. Andrews J
    66. Berger B
    67. Brenner SE
    68. Brent MR
    69. Cherbas L
    70. Elgin SC
    71. Gingeras TR
    72. Grossman R
    73. Hoskins RA
    74. Kaufman TC
    75. Kent W
    76. Kuroda MI
    77. Orr-Weaver T
    78. Perrimon N
    79. Pirrotta V
    80. Posakony JW
    81. Ren B
    82. Russell S
    83. Cherbas P
    84. Graveley BR
    85. Lewis S
    86. Micklem G
    87. Oliver B
    88. Park PJ
    89. Celniker SE
    90. Henikoff S
    91. Karpen GH
    92. Lai EC
    93. MacAlpine DM
    94. Stein LD
    95. White KP
    96. Kellis M
    97. mod EC
    98. modENCODE Consortium

    (2010) Identification of functional elements and regulatory circuits by Drosophila modENCODE

    Science 330:1787–1797.

    https://doi.org/10.1126/science.1198374

    • PubMed
    • Google Scholar
45. 1. Schubert T
    2. Pusch MC
    3. Diermeier S
    4. Benes V
    5. Kremmer E
    6. Imhof A
    7. Längst G

    (2012) Df31 protein and snoRNAs maintain accessible higher-order structures of chromatin

    Molecular Cell 48:434–444.

    https://doi.org/10.1016/j.molcel.2012.08.021

    • PubMed
    • Google Scholar
46. 1. Sigova AA
    2. Abraham BJ
    3. Ji X
    4. Molinie B
    5. Hannett NM
    6. Guo YE
    7. Jangi M
    8. Giallourakis CC
    9. Sharp PA
    10. Young RA

    (2015) Transcription factor trapping by RNA in gene regulatory elements

    Science 350:978–981.

    https://doi.org/10.1126/science.aad3346

    • PubMed
    • Google Scholar
47. 1. Simon MD
    2. Wang CI
    3. Kharchenko PV
    4. West JA
    5. Chapman BA
    6. Alekseyenko AA
    7. Borowsky ML
    8. Kuroda MI
    9. Kingston RE

    (2011) The genomic binding sites of a noncoding RNA

    PNAS 108:20497–20502.

    https://doi.org/10.1073/pnas.1113536108

    • PubMed
    • Google Scholar
48. 1. Sridhar B
    2. Rivas-Astroza M
    3. Nguyen TC
    4. Chen W
    5. Yan Z
    6. Cao X
    7. Hebert L
    8. Zhong S

    (2017) Systematic mapping of RNA-chromatin interactions in vivo

    Current Biology 27:602–609.

    https://doi.org/10.1016/j.cub.2017.01.011

    • PubMed
    • Google Scholar
49. 1. Ulianov SV
    2. Khrameeva EE
    3. Gavrilov AA
    4. Flyamer IM
    5. Kos P
    6. Mikhaleva EA
    7. Penin AA
    8. Logacheva MD
    9. Imakaev MV
    10. Chertovich A
    11. Gelfand MS
    12. Shevelyov YY
    13. Razin SV

    (2016) Active chromatin and transcription play a key role in chromosome partitioning into topologically associating domains

    Genome Research 26:70–84.

    https://doi.org/10.1101/gr.196006.115

    • PubMed
    • Google Scholar
50. 1. van Berkum NL
    2. Lieberman-Aiden E
    3. Williams L
    4. Imakaev M
    5. Gnirke A
    6. Mirny LA
    7. Dekker J
    8. Lander ES

    (2010) Hi-C: a method to study the three-dimensional architecture of genomes

    Journal of Visualized Experiments.

    https://doi.org/10.3791/1869

    • PubMed
    • Google Scholar
51. 1. Viollet S
    2. Fuchs RT
    3. Munafo DB
    4. Zhuang F
    5. Robb GB

    (2011) T4 RNA ligase 2 truncated active site mutants: improved tools for RNA analysis

    BMC Biotechnology 11:72.

    https://doi.org/10.1186/1472-6750-11-72

    • PubMed
    • Google Scholar
52. 1. Wang KC
    2. Yang YW
    3. Liu B
    4. Sanyal A
    5. Corces-Zimmerman R
    6. Chen Y
    7. Lajoie BR
    8. Protacio A
    9. Flynn RA
    10. Gupta RA
    11. Wysocka J
    12. Lei M
    13. Dekker J
    14. Helms JA
    15. Chang HY

    (2011) A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression

    Nature 472:120–124.

    https://doi.org/10.1038/nature09819

    • PubMed
    • Google Scholar
53. 1. Werner MS
    2. Ruthenburg AJ

    (2015) Nuclear fractionation reveals thousands of chromatin-tethered noncoding rnas adjacent to active genes

    Cell Reports 12:1089–1098.

    https://doi.org/10.1016/j.celrep.2015.07.033

    • PubMed
    • Google Scholar
54. 1. Will CL
    2. Lührmann R

    (2011) Spliceosome structure and function

    Cold Spring Harbor Perspectives in Biology 3:a003707.

    https://doi.org/10.1101/cshperspect.a003707

    • PubMed
    • Google Scholar
55. 1. Zhang Y
    2. Liu T
    3. Meyer CA
    4. Eeckhoute J
    5. Johnson DS
    6. Bernstein BE
    7. Nusbaum C
    8. Myers RM
    9. Brown M
    10. Li W
    11. Liu XS

    (2008) Model-based analysis of ChIP-Seq (MACS)

    Genome Biology 9:R137.

    https://doi.org/10.1186/gb-2008-9-9-r137

    • PubMed
    • Google Scholar
56. 1. Zhou Z
    2. Licklider LJ
    3. Gygi SP
    4. Reed R

    (2002) Comprehensive proteomic analysis of the human spliceosome

    Nature 419:182–185.

    https://doi.org/10.1038/nature01031

    • PubMed
    • Google Scholar

Author details

1. Jason C Bell

   Department of Biochemistry, Stanford University, Stanford, United States

   Present address

   Molecular Biology, 10x Genomics, Pleasanton, United States

   Contribution

   Conceptualization, Resources, Software, Formal analysis, Funding acquisition, Investigation, Methodology, Writing—original draft, Writing—review and editing

   For correspondence

   jason.bell@10xgenomics.com

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5480-7975

2. David Jukam

   Department of Biology, Stanford University, Stanford, United States

   Contribution

   Conceptualization, Resources, Software, Formal analysis, Funding acquisition, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4167-2754

3. Nicole A Teran

   1. Department of Biochemistry, Stanford University, Stanford, United States
   2. Department of Genetics, Stanford University, Stanford, United States

   Contribution

   Software, Formal analysis, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9625-5010

4. Viviana I Risca

   Department of Genetics, Stanford University, Stanford, United States

   Contribution

   Conceptualization, Resources, Software, Formal analysis, Funding acquisition, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2670-8704

5. Owen K Smith

   1. Department of Biochemistry, Stanford University, Stanford, United States
   2. Department of Chemical and Systems Biology, Stanford University, Stanford, United States

   Contribution

   Formal analysis, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0880-2801

6. Whitney L Johnson

   Department of Biochemistry, Stanford University, Stanford, United States

   Contribution

   Conceptualization, Resources, Funding acquisition, Investigation, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

7. Jan M Skotheim

   Department of Biology, Stanford University, Stanford, United States

   Contribution

   Resources, Funding acquisition, Writing—review and editing

   Competing interests

   No competing interests declared

8. William James Greenleaf

   1. Department of Genetics, Stanford University, Stanford, United States
   2. Department of Applied Physics, Stanford University, Stanford, United States

   Contribution

   Resources, Formal analysis, Funding acquisition, Methodology, Writing—review and editing

   Competing interests

   No competing interests declared

9. Aaron F Straight

   1. Department of Biochemistry, Stanford University, Stanford, United States
   2. Department of Chemical and Systems Biology, Stanford University, Stanford, United States

   Contribution

   Conceptualization, Resources, Formal analysis, Supervision, Funding acquisition, Methodology, Project administration, Writing—review and editing

   For correspondence

   astraigh@stanford.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-5885-7881

• 17,180

  views

• 2,312

  downloads

• 131

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.