Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Aug 27.
Published in final edited form as: Cell. 2015 Aug 20;162(5):1051–1065. doi: 10.1016/j.cell.2015.07.048

Genetic Control of Chromatin States in Humans Involves Local and Distal Chromosomal Interactions

Fabian Grubert 1,7, Judith B Zaugg 1,2,7, Maya Kasowski 1,7, Oana Ursu 1,7, Damek V Spacek 1, Alicia R Martin 1, Peyton Greenside 3, Rohith Srivas 1, Doug H Phanstiel 1, Aleksandra Pekowska 2, Nastaran Heidari 1, Ghia Euskirchen 1, Wolfgang Huber 2, Jonathan K Pritchard 1,4,5, Carlos D Bustamante 1, Lars M Steinmetz 1,2, Anshul Kundaje 1,6, Michael Snyder 1,*
PMCID: PMC4556133  NIHMSID: NIHMS717450  PMID: 26300125

SUMMARY

Deciphering the impact of genetic variants on gene regulation is fundamental to understanding human disease. Although gene regulation often involves long-range interactions, it is unknown to what extent non-coding genetic variants influence distal molecular phenotypes. Here, we integrate chromatin profiling for three histone marks in lymphoblastoid cell lines (LCLs) from 75 sequenced individuals with LCL-specific Hi-C and ChIA-PET-based chromatin contact maps to uncover one of the largest collections of local and distal histone quantitative trait loci (hQTLs). Distal QTLs are enriched within topologically associated domains and exhibit largely concordant variation of chromatin state coordinated by proximal and distal non-coding genetic variants. Histone QTLs are enriched for common variants associated with autoimmune diseases and enable identification of putative target genes of disease-associated variants from genome-wide association studies. These analyses provide insights into how genetic variation can affect human disease phenotypes by coordinated changes in chromatin at interacting regulatory elements.

INTRODUCTION

Deciphering the genetic and molecular basis of human traits and disease is a fundamental problem in biology and personalized medicine. Genome-wide association studies (GWAS) and deep sequencing efforts have identified common and rare single nucleotide genetic variants and structural variants associated with diseases ranging from inflammatory bowel disease and Alzheimer’s disease to cancer (Hindorff et al., 2009; Lambert et al., 2013; Rivas et al., 2011; Zuk et al., 2014). A majority of disease-associated common variants lie in poorly annotated non-coding genomic regions. Identifying target genes of these variants is complicated by the fact that many of these variants lie in distal enhancers (Harismendy et al., 2011; Maurano et al., 2012; Pomerantz et al., 2009; Schaub et al., 2012; Smemo et al., 2014), which can regulate genes several megabases (MBs) away through long-range chromatin contacts (Kleinjan and van Heyningen, 2005; Nobrega et al., 2003; Sanyal et al., 2012; Visel et al., 2009).

In this study, we integrated long-range chromatin contact maps with genetic variation and multiple molecular phenotypes to elucidate a comprehensive regulatory network of local and distal-acting regulatory and expression QTLs. We mapped inter-individual variation of three histone marks associated with enhancers and promoters in lymphoblastoid cell lines (LCLs) from a cohort of 75 individuals and used these to identify histone QTLs (hQTLs). In addition, we generated Hi-C and ChIA-PET chromatin contact maps (Fullwood et al., 2009; Lieberman-Aiden et al., 2009) and used them to identify genetic variants that act over large genomic distances through long-range interactions. We identified hQTLs for about 10% of predicted regulatory elements in LCLs. 15% of those were also associated with chromatin state changes at distal elements. Our findings indicate that genetic variation in regulatory elements can act over large distances to affect concordant changes in chromatin states and expression at distal sites often through TF motif disruptions. hQTLs in LCLs were enriched in GWAS SNPs associated with immune-mediated diseases allowing us to predict putative target regulatory elements and genes of several disease-associated common variants. Overall, our study provides novel insights into how genetic variation can affect human disease phenotypes by disrupting regulatory networks mediated by long-range chromatin interactions.

RESULTS

Profiling chromatin variation in a large cohort of lymphoblastoid lines enables comprehensive mapping of local histone QTLs

To study the relationship between genetic variation and chromatin activity, we generated ChIP-Seq data for three histone marks (H3K4me3, H3K4me1 and H3K27ac) in lymphoblastoid cell-lines (LCLs) from 75 unrelated individuals of the Yoruba (YRI) population, for whom high quality genotypes were available (1000 Genomes Project). H3K4me3 is primarily associated with promoters, H3K4me1 with active, bivalent, and weak enhancers, and H3K27ac with active promoters and enhancers (Ernst et al., 2011). We obtained ∼25 million uniquely mapping paired-end reads (2×101bp) per experiment (Table S1, Fig. S1A–B; Extended Experimental Procedures, “EEP”). In addition we utilized DNase-I hypersensitivity (DHS) data, which assays chromatin accessibility (for 68 YRI individuals (Degner et al., 2012)), and RNA-Seq data to quantify gene expression (for 54 YRI individuals (Lappalainen et al., 2013)).

To identify single nucleotide polymorphisms (SNPs) associated with chromatin variation, we performed quantitative trait locus (QTL) mapping. We used the ChIP-seq and DHS data to identify and quantify normalized signal at regions of enrichment (peaks). We also quantified normalized expression levels for all genes using the RNA-seq data (EEP, Fig. S1C–D). All signal measurements were corrected for confounding factors (EEP, Fig. S1E–F). To identify local QTLs, we tested for associations of histone/DHS peaks/gene expression and the most strongly associated SNP within +/−2kb of peak boundaries/promoters using a linear regression framework (Degner et al., 2012), resulting in a total of 22,624, 14,142, and 9,575 local hQTLs (10% FDR) linked to 10–15% of H3K4me1, H3K27ac and H3K4me3 peaks respectively, as well as 2,450 dsQTLs and 933 eQTLs (Fig. 1A–B, Fig. S1G, Table S2; EEP).

Figure 1. Local QTLs.

Figure 1

Open in a new tab

A) Number of local QTLs (10% FDR) for histone marks, RNA expression (Lappalainen et al., 2013) and DHS sites (Degner et al., 2012). Peaks are classified as promoters (“TSS”), “enhancer (Enh)” and “other” based on chromatin states (Kasowski et al., 2013).

B)Chromosome-wide distribution of local QTLs on chromosome 1 for histone marks, DHSs and RNA. The vertical red line marks the location of the local joint QTL described in panels C-E.

C) Signal tracks for three histone marks and DHS in a 5kb region around a joint local hQTL/dsQTL/eQTL coinciding with the ZNF695 promoter. The signal is aggregated across individuals by genotype at rs61373194. The position of the QTL SNP is indicated with a vertical dashed red line.

D) Boxplots of aggregated signal for gene expression, histone marks and DHS grouped by the genotype of the QTL SNP. The normalized signal corresponds to the signal averaged across the entire peak region as indicated by black dashed lines in panel C.

E) Position weight matrix for SPI1. The fourth position corresponds to the location of the motif altering SNP (rs61373194). The genotype with the strongest signal corresponds to a better match to the consensus sequence. Note: the motif is on the (-) strand. Therefore the genotypes are shown for the (-) strand to correspond to the PMW.

F) Overlap between local eQTLs and hQTLs (calculated for promoters within 5kb of a histone peak). The overlap is highly significant (Fisher’s exact test); two thirds of all eQTLs are also hQTLs. There are 8,239 promoters that coincide with non-QTL histone peaks.

74% and 19% of all local QTLs were associated with enhancer and promoter chromatin states, respectively (EEP). QTLs often jointly influence multiple molecular phenotypes within the same region (Fig. S1H). One example is shown in Fig. 1C–E where disruption of the SPI1 (PU.1) pioneer TF binding site is associated with decreased signal of three histone marks (hQTL), chromatin accessibility (dsQTL), and RNA expression (eQTL) at the ZNF695 gene promoter (Fig. 1C–E). Globally we find that 66% of local eQTLs are also hQTLs (3.5-fold enriched, p-value=2.2e-16, Fisher’s exact test; Fig. 1F; EEP). eQTLs that do not coincide with hQTLs may influence RNA processing rather than steady-state gene expression (Lappalainen et al., 2013), as over 30% of them reside in the 5’UTR of at least one mRNA isoform (EEP).

Long-range chromatin contacts are associated with concordant regulatory variation across distal genomic elements

Long-range chromatin contacts are responsible for physical interactions between distal regulatory and transcribed elements. We hypothesized that genetic variants associated with local variation in chromatin state may also correlate with the activity of physically linked distal elements. To address this, we generated Hi-C data (1.4 billion reads) in a reference LCL (GM12878) to obtain a chromatin contact matrix (resolution at single restriction fragment length, median = 2,274 bp, mean = 3,697 bp). We used a covariance measure as a robust proximity estimate to identify putatively physically interacting elements (proximity score > 0.4; Fig. S2A–E; EEP). We tested whether a SNP is more likely to affect a coordinated gain or loss of histone modification, DHS, or RNA signal at pairs of local (<2kb from SNP) and distal (> 50kb) elements if the pairs showed evidence of physical interaction (Fig. 2A). Consistent with our hypothesis, pairs of physically interacting peaks, or genes, showed more concordant associations of SNPs with local and distal chromatin variation (Fig. 2B,C, Fig. S2F).

Figure 2. Evidence of Genetic Coordination Among Distal Functional Elements.

Figure 2

Open in a new tab

A) Schematic of the four possibilities for genetic and physical coordination: the effect of a local QTL on a distal peak can be in the same (+/+ and −/−) or opposite direction (+/− and −/+) and the local QTL can be physically interacting with distal peak or not (bottom vs. top).

B) The effect (beta of the regression) of a local QTL on its local peak (y-axis) and distal H3K27ac peak (x-axis) grouped by presence (“interacting”) or absence (“non interacting”) of Hi-C physical links (bottom vs. top). The direction of the effect tends to in the same direction for physically interacting pairs but not for non-interacting pairs (see also Fig. S2F for QTLs distal to the other marks, DHSs, and RNA).

C) Enrichment of physical interaction among pairs of local QTLs and all distal genomic features (>50kb) that covary in the same direction with respect to the local QTL SNP. The enrichment is shown for all combinations of histone marks, RNA, and DHSs. (Fisher’s exact test, red bars indicate significance with *p<0.01 and **p<10−10).

D) Enrichment for genetic associations (p-value < 10−6) in physically interacting regions as a function of the minimal distance between the SNP-distal peak pair. Enrichment and confidence intervals are calculated using Fisher’s exact test.

Since regions that are proximal in linear genomic distance are also more likely to physically interact, Hi-C supported SNP-peak pairs span shorter genomic distances than non-interacting pairs. To account for this distance bias, we calculated the enrichment of Hi-C interactions linking local-distal pairs that showed a genetic association (p-value < 1e-6) for increasing distance cut-offs (EEP). As distance increases between a SNP and distal element, Hi-C linked pairs are increasingly enriched for genetic associations (Fig. 2D). The effect is particularly strong for RNA expression (>10-fold enrichment of genetic associations in Hi-C connected fragments at distances >200kb). Interestingly, physically interacting regions involving distal DHS peaks are the exception to this phenomenon (Fig. 2C, Fig. S2F). This could be explained by the largely local nature of DHS regulation in contrast to more distally coordinated regulatory changes in histone marks, or alternatively it could simply be due to the greater power of our deeply sequenced histone data.

Long-range chromatin contact maps increase power for distal QTL discovery

Given the support for genetically coordinated regulatory variation through long-range chromatin contacts, we next sought to identify local QTLs associated with physically interacting distal (>50kb) molecular phenotypes (“distal QTLs”). We paired SNPs that were local h/ds/eQTLs with distal (within 2Mb) histone/DHS peaks and genes, and grouped the pairs into whether or not they are physically interacting based on Hi-C proximity scores. We then assessed the significance of the associations independently for both groups, leveraging the Hi-C data as an orthogonal filter (Bourgon et al., 2010). To ensure that our distal QTLs were independent (i.e. not in linkage-disequilibrium (LD) with a local QTL for the same peak) we (i) identified and kept the most significant SNP (regardless of distance) for each peak; (ii) removed all SNPs in LD (r2>0.2) with the SNP identified in the previous step; repeated steps (i) and (ii) until no SNPs were remaining (see EEPs, Fig. S3D). From this set of independent QTLs for each peak/gene, we defined distal QTLs as being located >50kb from the peak/gene (Table S2).

The resulting QTLs were compared to those obtained from a standard QTL analysis. Using Hi-C as a filter we identified 20,950 distal QTLs across all tested molecular phenotypes as compared to 8,305 using the standard approach (10% FDR; Fig. 3A. Fig. S3A, EEP). Hi-C data increases power to detect distal QTLs irrespective of genomic distance constraints with increased gains for larger window sizes by focusing the association tests on SNP-peak pairs that are more likely to be associated (Fig. 3B; control of permuted Hi-C links Fig. S3B; Fig. S3C; EEP). Most distal QTLs were located within 200 kb from the associated functional element (Fig. 3C). We hypothesized that if QTLs can affect regulatory variation at distal elements through transient chromatin contacts, it is likely that a variant will have a weaker effect on a distal element compared to a local element. Consistent with this hypothesis, we find smaller effect sizes for distal QTLs compared to local QTLs (p-values <10−16 (Wilcoxon rank sum test); Fig. 3D).

Figure 3. Distal QTLs.

Figure 3

Open in a new tab

A) Number of distal QTLs identified with and without Hi-C. To identify distal QTLs we used SNPs that are a local QTL for any of the histone marks or RNA, split the set of SNP-peak pairs based on whether or not they are physically interacting (Hi-C correlation > 0.4), and calculated the FDR for each set individually (see EEP and Fig. S3A–D).

B) Q-Q plot for the different sets of SNP-peak pairs for H3K27ac. The expected p-values were calculated by permuting sample labels (see EEP). Q-Q plots for the other marks as well as a control set using permuted Hi-C links is shown in Fig. S3A–B.

C) Distance distribution for local-distal hQTL peak pairs identified using Hi-C interaction data and those identified using a standard hQTL approach. Most distal QTLs are less than 200kb apart from their targets.

D) Distribution of absolute effect sizes of local and distal QTLs. Globally, local effects are stronger than the distal effects.

E) Heat map of estimated Hi-C interaction counts for a region harboring multiple significant distal and local QTLs. The Hi-C interactions are based on a covariance method (EEP). Hi-C fragments with a correlation score > 0.4 were considered interacting. Local QTLs are indicated on the diagonal and off-diagonal dots indicate distal QTLs (>50kb). Black squares correspond to contact domain calls from (Rao et al., 2014).

F) The position of distal QTLs was mirrored (see schematic) to test whether local-distal QTL pairs are more likely to reside within the same TAD. The schematic indicates how QTL SNPs were mirrored relative to the associated feature (histone peak, DHS, or RNA). Shown is the fraction of feature-QTL pairs that share the same TAD (orange) compared to feature-mirrored QTL pairs (grey). Features more frequently share the same TAD with the true QTL than the mirrored QTL. To avoid any bias we used the non-Hi-C aware set of QTLs for this analysis.

G) Distribution of fractions of local-distal H3K4me1 QTLs sharing the same TAD for 100 sets of shuffled TADs. Real data corresponding to true TADs is indicated by the red vertical line. For chromatin marks, the fraction of local-distal hQTLs that share the same TAD is significantly higher than for the shuffled TADs (One-sided t-test). To avoid any bias we used the non-Hi-C aware set of QTLs for this analysis. The same analysis for the other marks is shown in Fig. S5D.

H) Comparison of replication timing for local-distal QTL pairs. Shown are scatterplots of the replication timing for a H3K27ac peak region (y-axis) and the region of its distal QTL (x-axis; right panel) or mirrored QTL (see schematic in Fig. 4F) (x-axis; left panel). The peak-real QTL pairs show higher correlation than the peak-mirrored QTL pairs (Pearson correlation).

I) The cumulative distribution of replication timing difference between peak and QTL regions (orange) and peak and mirrored QTL regions (grey) is shown. The differences for the hQTL- real peak region pairs are significantly smaller than for the corresponding mirrored positions (Wilcoxon rank sum test).

Distal chromatin QTLs are enriched within topologically associated domains

Distal QTLs for multiple molecular phenotypes were found to often fall in regions with high density of physical interactions (Fig. 3E) which highly overlap with topologically-associated domains (TADs)(Dixon et al., 2012; Rao et al., 2014) (Fig. S2E). Mirroring the real position of the SNP relative to the distal molecular trait produced fewer pairs residing within TADs (Fig 3F), suggesting that distal chromatin QTLs are likely to be enriched within TADs. To explicitly test this hypothesis, we shuffled the locations of TADs and compared the number of distal QTLs falling within the actual TADs versus numerous shuffled sets (EEP). We observed a significant enrichment (p-values <10−70; one-sided t-test) of distal QTLs involving all three histone marks (Fig. 3G, Fig. S3E) and a corresponding lack of enrichment for distal associations across TAD boundaries (Fig. S3F). This result was confirmed in an alternate analysis in which we examined the fraction of distal QTLs falling within TADs compared to random sets of distance-matched SNP-peak pairs. Again, distal QTLs for all three histone marks, but not RNA (p=0.19) and DHS (p=0.25), were found to be enriched within TADs (p-values ≤0.01; Fisher’s Exact Test). Similarly, the replication timing (Koren et al., 2012) of pairs of regions connected by a distal QTL were more similar compared to equi-distant control regions showing no significant genetic associations (Fig. 3H–I).

hQTLs form a highly connected network of diverse physically interacting regulatory and transcribed elements

To further dissect the functional elements associated with regulatory variants, we utilized combinatorial chromatin states (Kasowski et al., 2013) and GENCODE v19 transcriptome annotations (Harrow et al., 2012) to label peaks as putative enhancers or promoters (EEP). The majority of local-distal QTLs (88%) capture associations between and across enhancers and promoters (Fig. 4A). QTLs involving enhancer-enhancer pairs represent the largest and, for the enhancer marks (H3K4me1, H3K27ac) a significantly higher than expected, proportion (54%) of distal hQTLs (Fig. 4A,B). 5.7% of QTLs involve pairs of local and distal promoters (>2-fold more than expected for all marks and RNA; Fig. 4B) consistent with Pol II ChIA-PET studies reporting extensive promoter-promoter links (Li et al., 2012).

Figure 4. Characterization of Distal QTLs.

Figure 4

Open in a new tab

A) Fraction of chromatin states for local-distal QTL pairs grouped by mark at distal site. Most pairs involve associations between enhancers and enhancers (yellow) or enhancers and TSSs (orange).

B) The log odds ratio of observed vs. expected combinations of chromatin states for local and distal QTL pairs is shown for each chromatin mark, RNA and DHS. The number of expected combinations for each mark was obtained by permuting the state labels for the local peaks. (Fishers exact test * = p-value < 0.01)

C+D) Number of local enhancer QTLs (C) and local promoter QTLs (D) influencing one or more distal promoters or enhancers (“out-degree”). In-degree plots are shown in Fig. S4.

E) Out-degree (as in Figure 4BC) of local enhancer QTLs that are also eQTLs for one or more distal genes.

F) In-degree of genes with one or more distal enhancer/promoter.

G) Pearson correlation of expression signal among pairs of distal promoters that are associated with the same hQTL (one local, one distal). The correlations are compared to a set of distance matched permuted pairs (permuted genes) and a set of real links with permuted sample names (permuted individuals). Promoters sharing a hQTL are more correlated than permuted data (Wilcoxon rank sum test, p=2.8e-12).

We systematically studied the properties of genetically coordinated regulatory elements. For enhancer and promoter local hQTLs, we computed the “out-degree”, i.e., the number of outgoing associations to distal enhancers and/or promoters (Fig. 4C–D). For regulatory elements that are linked by at least one distal QTL, we computed their number of associated incoming distal hQTLs (“in-degree”; Fig. S4A–C). QTLs local to enhancers or promoters are frequently associated with one or more distal enhancers (3,934) and promoters (1,384) (Fig. 4C–D). We identified 184 local enhancer hQTLs with distal effects on gene expression, 12.5% of them associated with more than 1 gene (Fig. 4E). Reciprocally, 26% of the 151 genes with an eQTL are associated with SNPs at multiple enhancers. (Fig. 4F). Interestingly, we identified 84 promoter hQTLs that are eQTLs for a distal gene (Fig. 4F). As expected, these pairs of genes with a shared regulatory variant showed significantly more correlated expression than distance-matched random pairs of genes (Fig. 4G; p-value=1.5e-6, Wilcoxon rank sum test). Overall, these results suggest extensive overlap of eQTLs and hQTLs, and that a subset of joint hQTL-eQTLs involve multiple regulatory elements or genes.

Looping chromatin contacts connect covarying loci that are enriched for genetic associations

To independently assess the extent of genetic associations between physically interacting loci, we generated ChIA-PET data for the H3K4me3 promoter mark and RAD21 (a subunit of the cohesin complex involved in mediating distal chromatin contacts) in GM12878 cells (Fig. 5A; EEP). We identified 41,532 chromatin loops for RAD21 and 10,220 for H3K4me3 (5% FDR), a large portion (∼80%) were found to overlap with loops uncovered in a previous study (Rao et al., 2014) indicating the good quality of our data (Fig. S5A). We characterized the loops by chromatin state at interacting sites. As expected (Jin et al., 2013; Rao et al., 2014), interactions between enhancers and promoters were highly prevalent (9,147; 24% of all ChIA-PET links), often involving multiple enhancers (2,085) linked to the same promoter (Fig. 5B; EEP). We also find ∼2,028 promoters (5% of ChIA-PET links) linked to a different distal promoter, and 470 promoters linked to multiple distinct promoters (Fig. 5B). H3K4me3 ChIA-PET was enriched for promoter-enhancer interactions whereas the RAD21 ChIA-PET was enriched for promoter-promoter interactions and “other” chromatin state classes including repressed and CTCF states (Fig. 5C; Fisher’s exact test).

Figure 5. Chromatin Loops and ChIA-PET.

Figure 5

Open in a new tab

A) Example locus showing a genetic variant (rs4405472; dashed vertical line) that is a local hQTL for H3K4me1 and H3K4me3 as well as a distal hQTL for H3K4me3 at a promoter ∼ 200kb away, skipping several genes. The two loci are physically linked as indicated by significant ChIA-PET interaction calls (top) and a Hi-C based chromatin loop call (bottom, (Rao et al., 2014). The green line indicates a chromatin contact domain (i.e. TAD).

B)ChIA-PET-links between one or more enhancers/promoters and distal genes based on interaction calls for H3K4me3 and Rad21.

C) The log odds ratio of observed vs. expected state combinations for ChIA-PET linked peaks. Expected state combinations are calculated by exhaustively counting all possible combinations of states present at ChIA-PET peaks (Fisher’s Exact test, red=p<0.01)

D) Enrichment of genetically associated (p value <10−6) local-distal hQTL pairs in interacting ChIA-PET fragments (Rad21) or chromatin loops as defined by Rao et al. for each histone mark and RNA. The 5% and 95% confidence intervals are shown in black (*p<0.01; Fisher’s exact test).

E) Pairs of promoters that are physically linked (ChIA-PET) show correlation of signal for several features. All features except DHS show significantly higher correlation than a control of permuted, distance-matched links or permuted sample labels (Wilcoxon rank sum test)

F) Same as G) but for enhancers linked by ChIA-PET. Interacting enhancers are more correlated than any of the permuted, distance-matched data (p value <10−16 for all histone marks and DHS; Wilcoxon rank sum test).

We further used our ChIA-PET data along with the Hi-C data from Rao et al. to test if physically interacting loci are enriched for associations between SNPs and distal molecular phenotypes. We observed a significant enrichment of both ChIA-PET and Hi-C loops in distal associations (P < 10−6) between SNPs and all three histone marks (not significant for H3K4me3 in Hi-C loops), as well as RNA expression levels (Fig 5D, EEP). This result was confirmed in an alternate analysis in which we shuffled the location of the loops and counted the number of associations involving SNP-peak pairs located on opposite ends of chromatin loops relative to multiple sets of shuffled loops (EEP). The ChIA-PET and Hi-C loops were both enriched in associations between a SNP and two of the three histone marks (H3K4me3 and H3K27ac), and RNA expression levels (Fig. S5B), with expression showing the strongest enrichment (∼2.5-fold and ∼2.2-fold, in ChIA-PET and Hi-C loops, respectively).

To corroborate our observation that local-distal hQTLs tend to physically interact, we tested whether pairs of regulatory elements linked by ChIA-PET interactions also exhibit concordant variation of molecular phenotypes. Indeed, we observed significant covariation of histone/DHS signal and expression levels (Fig. 5E–F; Wilcoxon rank sum test p<1e-16) amongst pairs of enhancers/promoters and genes linked by ChIA-PET interactions.

hQTLs potentially affect chromatin state at regulatory elements by disrupting binding sites of transcription factors

To investigate the potential molecular basis of hQTLs, we used ENCODE TF ChIP-seq data from GM12878 (Dunham et al., 2012) to assess the enrichment of in-vivo TF binding sites at regulatory elements associated with local hQTLs. For each TF, we computed the enrichment of overlap between histone peaks containing at least one TF peak and those associated with a local hQTL (Fisher’s exact test). While hQTL associated H3K27ac peaks were enriched for key lymphoid TFs such as SPI1 (PU.1), BCL11A and PAX5 hQTL associated H3K4me3 peaks were enriched for general regulatory factors such as POLR2A, USF1, ELF1 and TAF1 (Fig. 6A–B, Fig S6A–B). A complementary, rank-enrichment approach (Fig. 6C) (EEP) revealed similar results: the highest scoring TFs by the Fisher’s exact test (e.g. SPI1) show significant overlap enrichments across the entire spectrum of histone ChIP-seq peaks; low scoring TFs (e.g. NRF1 and FOS) exhibit a distinct lack of enrichment at peaks showing significant QTL associations (Fig. 6C).

Figure 6. Transcription Factor Binding Analysis at Local and Distal hQTLs.

Figure 6

Open in a new tab

A) Overlap enrichment of TF binding in H3K27ac peaks that have a QTL. For each TF, we plotted enrichment of having a QTL and being bound by the respective TF. Bars represent the 95% confidence interval. In red are significant enrichments, in gray non-significant ones.

B) Overlap enrichment of TF binding in histone mark peaks, focusing on peaks with a local QTL (first three columns) or peaks affecting distal sites (last three columns). The enrichment value is only plotted for TF-histone mark pairs for which the enrichment was significant. Rows are sorted by the fold enrichment for peaks with a local H3K27ac QTL, which is displayed in A).

C) Rank enrichment of TF binding in H3K27ac peaks. We plot the fold change enrichment of H3K27ac peaks in TF binding sites at increasing levels of significance for called hQTL peaks (red). The background enrichment was obtained by permuting the p-values between peaks (gray).

D) Number of total peaks that can be explained through correlation between a TF motif score and the molecular phenotype. For each molecular phenotype, we computed the correlation between signal and motif score, and defined significant motif disruptions using permutations (5% FDR). The number of peaks with at least one significantly correlated TF motif within 2kb is shown. For each molecular phenotype, peaks were grouped by the number of motif-disrupting SNPs per peak.

E) Fraction of H3K27ac hQTLs that are significantly correlated with TF motif disruptions. TF motifs show positive and negative correlation with the local histone mark signal and are sorted by the difference between percent positive and negative correlations. The total number of tested SNP-peak pairs across all H3K27ac peaks is annotated next to the TF name. Only TFs with N>=50 are shown.

F) The signal surrounding H3K27ac QTNs was extracted and grouped into 6 clusters (pam clustering, EEP). The aggregate signals for the 6 clusters are shown for the high-, heterozygous- and low-genotypes (blue, purple, red) for H3K27ac, H3K4me3, H3K4me1, and DHS. Nucleosome positioning is indicated by MNase signal extracted from the same regions for a single individual (left to right; signal heat maps are shown in Fig. S6H). As expected, histone signal coincides with MNase signal / nucleosomes, whereas DNase hypersensitivity coincides with nucleosome-free regions. QTNs show concordant effects on the three histone marks.

We next tested whether disruption of TF binding motifs could explain the observed variation of histone marks. For each TF, we used one representative motif position weight matrix (PWM) (Kheradpour and Kellis, 2014). After extensive filtering to avoid false positive calls of motif matches across the genome (EEP), we computed the Spearman correlation between the motif PWM score at the SNP and the histone mark signal at the peak, across all individuals. For the subset of peaks/genes with a QTL that passed filtering (13% H3K4me1, 25.9% H3K4me3, 18.7% H3K27ac, 33.3% DHS and 32.4% RNA), we found significantly correlated motifs for 89–95% of peaks (5% FDR; Fig. 6D, see Table S3 for full analysis results). For H3K27ac, TFs with the highest proportion of positive correlations include activating enhancer-associated TFs (e.g. NFKB, BCL1, SPI1), whereas TFs with known repressive roles (e.g. REST, ZBTB33, SRF) show a larger proportion of negative correlations (Fig. 6E, see Fig. S6C–F for all histone marks, DHS and RNA). Similarly, we found a correlated motif for the majority of tested distal QTL peaks (60–73% for histone marks and DHS, 35% for RNA). Joint analysis of motif correlations with local and distal peaks revealed largely concordant effects (Fig. S6G), consistent with co-variation of the distally associated peaks (Fig. 2B).

To further characterize the local chromatin environment of motif-disrupting quantitative trait nucleotides (QTNs), we extracted ChIP-seq signal for H3K27ac around each H3K27ac QTN (± 500 bp), averaged the signal across individuals at each location, sorted the profiles into six groups using unsupervised clustering, and aggregated the signal in each cluster by genotype (EEP Fig. 6F). Consistent with TF binding being a major driver of chromatin variation, we observe that the vast majority of QTNs lie in nucleosome-free regions, either between two peaks (clusters 1,2,3) or at peak shoulders (cluster 5,6) with only a small subset potentially coinciding with nucleosomes (cluster 4) (Fig. 6F). Grouping and aggregating the signal for other histone marks and DHS of the same regions according to the H3K27ac clusters, we found high concordance between histone marks and DHSs (Fig. 6F, S6F). MNase-seq data from GM12878 (Kundaje et al., 2012) confirmed that the chromatin mark patterns are consistent with overall nucleosome positioning (Fig. 6F, S6F).

hQTLs are enriched for autoimmune disease-associated variants enabling the identification of putative disease-associated target genes and regulatory networks

Enhancer elements active in disease-relevant cell types have been previously shown to be enriched for disease-associated variants from genome-wide association studies (GWAS) (Karczewski et al., 2013; Maurano et al., 2012; Nicolae et al., 2010; Schaub et al., 2012). Consistently, we found significantly more GWAS SNPs in strong LD (r2 ≥ 0.8) with hQTLs than SNPs matched for MAF, distance to promoter/TSS, and LD (Fig. 7A; data from Table S4). When separating hQTLs into enhancers and promoters, we found a significant enrichment of GWAS SNPs for both categories in all three marks (Fig. 7A, Fig S7A).

Figure 7. Effect of regulatory elements and hQTLs on phenotypic diversity.

Figure 7

Open in a new tab

A) Promoter and enhancer histone peaks with an hQTL are significantly enriched for GWAS SNPs compared to histone peaks without an hQTL. Dashed lines indicate the value for GWAS SNPs, while the null distributions indicate values by matching on GWAS SNPs for MAF, LD, and distance to TSS.

B) Enrichment of hQTLs in GWAS SNPs. We plot the negative log10 adjusted p-value by log2 odds ratio of the Fisher’s Exact test.

C–D) Rank enrichment of hQTLs in GWAS SNPs. At each tier of significance for GWAS variants we plot the fold change enrichment in overlap with hQTLs compared to base overlap of all GWAS variants at any significance level (red). We permute the p-values between the GWAS variants to generate background enrichments (gray). We select one positive enrichment and one negative enrichment example for each mark H3K27ac, H3K4me1 and H3K4me3.

E) Regulatory network for Crohn’s disease derived from hQTLs and eQTLs intersected with GWAS SNPs. The network consists of 1) GWAS tag SNPs (gray boxes) connected to QTL SNPs (white nodes) through edges representing LD (gray edges), and 2) QTL SNPs connected to affected regulatory elements (orange triangles) or genes (green nodes) through edges representing either local QTLs (solid lines) or distal QTLs (dotted lines). An orange edge represents an hQTL/dsQTL, a green edge represents an eQTL. Distal QTL edges are labelled with the distance between the QTL SNP and the midpoint of the regulatory element. Regulatory elements are defined as the merged peaks from H3K4me1, H3K4me3, H3K27ac and DHS. If the affected peak in a regulatory element overlaps the transcription start site of a gene, we label the regulatory element with the gene’s name. If multiple QTL SNPs are associated with the same local regulatory element, we label the QTL SNP with the rsID of the most significant SNP; if multiple QTL SNPs are associated with the same distal regulatory element, then for every combination of local-distal association we pick the best-correlated SNP for the local peak. (Here, we only show the parts of the network that contain a gene. For the full network see Figure S7.)

Further, we found that hQTL SNPs for all histone marks are significantly enriched (Fisher’s Exact test) for SNPs associated (association p-values < 1e-5) with autoimmune diseases including multiple sclerosis, rheumatoid arthritis, Crohn’s disease and ulcerative colitis (Fig. 7B). These findings were corroborated using rank enrichment analysis for the subset of diseases/traits for which complete summary statistics were available (Table S4) (EEP): hQTLs were enriched in GWAS SNPs with moderate to strong associations with autoimmune diseases (Fig. 7C–D, Fig. S7B), but were not enriched for coronary artery disease (Fig. S7B). Both methods also exhibit strong enrichment of LCL hQTLs in Alzheimer’s disease (AD) associated GWAS variants (Fig. 7B,7D), consistent with recent findings indicating that AD GWAS variants are strongly enriched in regulatory elements active in CD19+ primary B cells and CD14+ primary monocytes (Gjoneska et al., 2015).

To identify putative target genes of disease-associated hQTLs, we intersected our local and distal (GWAS were expanded using European LD, r2>=0.8, QTLs were expanded using Yoruban LD, r2>=0.8) hQTLs and eQTLs with genome-wide significant GWAS tag SNPs (EEP) for each enriched disease (Table S4 and Fig. S7C). Our analysis recapitulated several known target genes and provided other new candidates. E.g. using Crohn’s Disease (CD) as a case study, we linked 115 of 425 genome-wide significant (p <10^-5) tag SNPs to 184 regulatory elements and 33 genes (Fig 7E). We find distal hQTLs/dsQTLs for SLC25A20, WDR6, APEH, RHOA and TCTA as well as distal eQTLs for RP11-387H17.4, ORMDL3 and ERAP2. eQTLs in LD with CD GWAS SNPs have been previously reported for the ERAP2 and ORMDL3 genes, (Franke et al., 2010; Van Limbergen et al., 2009).

DISCUSSION

We have integrated genetic, chromatin, and expression variation data with chromatin contact maps in LCLs to characterize the effects of genetic variants on distal regulatory elements and gene expression. We developed a novel approach using 3D chromatin contact maps from Hi-C data as a scaffold for long-range association testing that significantly increases power to detect distal hQTLs by restricting hypothesis testing to sites of physical proximity. Most importantly, our study provides mechanistic insights into genetically coordinated variation occurring through three-dimensional physical contacts of regulatory elements.

In utilizing 3D proximity to discover QTLs we make a number of observations pertaining to the role of chromatin structure in regulatory variation. We found that local and distal hQTLs are abundant, suggesting a strong genetic basis for the extensive variation of chromatin state (Kasowski et al., 2013). Consistent with the high frequency of physical interactions within topologically associated domains (TADs), we find that pairs of local-distal hQTLs are enriched within these domains and at looping chromatin contacts. Our results further suggest that specific genetic variants harbored within regulatory elements may concordantly affect local and distal histone modifications through interacting networks of elements. Such concordance could involve a variant in one element influencing another element in physical proximity by recruitment of transcriptional regulators. Alternatively, the process of transcription itself could influence local chromatin (e.g. either directly through polymerase mediated recruitment of chromatin modifying enzymes, or indirectly by affecting recruitment of histone modifying enzymes by a lncRNA whose expression is directly affected) or the QTL may influence chromatin state by affecting transcription of enhancer RNAs.

Our results hold special relevance for medical genomics. The robust disease enrichments we observe suggest that long-range histone QTL mapping in diverse cell types will help elucidate the relevant cell type specific functional elements that give rise to GWAS signal. Furthermore, we propose that hQTL mapping is an important tool for linking GWAS tag SNPs to putative target genes, regulatory elements, and networks bypassing the need to first identify the causal variant (Fig. 7). Future investigations of these regulatory networks anchored on GWAS variants may help elucidate the genetic and molecular mechanism underlying diseases and traits.

EXPERIMENTAL PROCEDURES

ChIP-Seq

Cells were cross-linked with formaldehyde at 1% final concentration for 10 minutes. Chromatin corresponding to 20 million cells was sheared and subjected to immunoprecipitation with antibodies to H3K4me3, H3K4me1 or H3K27ac, respectively. Enriched fragments were subjected to Illumina TruSeq library preparation and paired-end sequencing.

ChIP-Seq and RNAseq data processing

We aligned histone mark ChIP-seq datasets for H3K4me1, H3K4me3 and H3L27ac to personal genomes and subsampled the aligned reads to ensure consistent sequencing depth. We then called peaks on each individual using MACS (Zhang et al., 2008), merged peaks across all individuals, and extracted signal in each region for each individual. This was followed by quantile normalization, standardization and batch effects removal using PEER (Stegle et al., 2010). We used a similar strategy for the DNase data from (Degner et al., 2012). For quantifying RNA levels from GEUVADIS (Lappalainen et al., 2013), we used Sailfish (Patro et al., 2014), followed by the above normalization scheme on TPM values.

Hi-C

25 million cells for GM12878 were cross-linked, nuclei lysed and chromatin digested with HindIII. DNA overhangs were biotinylated and proximity ligated under dilute conditions to favor ligation of fragments in three-dimensional proximity. DNA was then sheared, and biotinylated fragments were pulled down with streptavidin beads to enrich for physically interacting sites. Libraries were prepared for Illumina paired-end sequencing and data was processed with the HiCUP pipeline (http://www.bioinformatics.babraham.ac.uk/projects/hicup/) to obtain the interaction counts for each pair of restriction fragments. Based on the assumption that two sites that largely interact with the same set of sites are likely to also interact with each other we calculate the proximity between two restriction fragments A and B as the ratio between fragments that interact with both A and B and the number of fragments interacting with only one of them. The threshold of 0.4 for calling an interaction significant was determined based on concordance measures across pseudo replicates (Fig. S2; EPP).

ChIA-PET

Cells for GM12878 were cross-linked and chromatin was prepared for immunoprecipitation with antibodies against Rad21 and H3K4me3, respectively. Immunocomplexes were pulled down with Protein-G dynabeads, marked with biotinylated linkers and subsequently proximity ligated. Biotinylated fragments were pulled down with Streptavidin Dynabeads and libraries prepared for Illumina paired-end sequencing. Physical interactions have been called using the Mango pipeline which corrects for known biases in ChIA-PET experiments, such as genomic distance (Phanstiel et al., 2015).

Local hQTL calling

For identifying the local histone QTLs we searched for the best correlated SNP within 2kb of the peak boundaries. The false discovery rate (FDR) was estimated based on an distribution of empirical p-values obtained from associations from 1000 sets of peak-wise independent permutations of the sample labels (Fig. S1; EPP).

Distal hQTL calling

We only tested SNPs that were associated with a local peak above the 10% FDR threshold for local associations. We then calculated associations for all SNP-peak pairs within 2MB. To correct for LD and obtain a set of independently associated SNPs for each peak we applied the following algorithm: (i) identify and keep the most significant SNP (regardless of distance) for each peak, (ii) remove any SNP in LD (R2>0.2), repeat (i) and (ii) until no SNP was left (see EEPs, Fig. S3D). We then split the set of SNP-peak pairs into HiC-interacting and non-interacting pairs and calculated the FDR independently for each set using the a permutation-based empirical p-value distribution (1000 sets of per-peak independent sample label permutations).

Enrichment in topological domains

To assess the enrichment of distal QTLs within TADs we (i) calculated the overlap relative to a set of shuffled domains and (ii) compared the fraction of same-domain QTL pairs to a set of control positions which were obtained by mirroring the position of the QTL relative to the associated distal peak (Fig. 3F,G).

TF motif analysis

We tested enrichment of TF binding sites in the peaks that have QTL using a Fisher’s Exact Test. Then, we checked whether TF motif disruption (computed as TF motif PWM scores computed at each SNP per individual) correlates with the signal at histone marks and DHS, or with RNA expression, using Spearman correlation. We assessed significance by permutation testing, with an FDR of 5%.

GWAS analysis

We overlapped hQTL SNPs for H3K27ac, H3K4me and H3K4me3 with GWAS SNPs with p<1e-5 for over 150 diseases. We compared the overlap to two negative sets: the set of GWAS SNPs with p>1e-5 and a set of SNPs matched for MAF, LD and distance to TSS. We assessed difference in enrichment with a one-sided Fisher’s exact test. We first thinned GWAS SNPs to keep only the most significant variant of those variants in EUR LD with r2>0.8 and then considered an overlap with any SNP in EUR LD r2>0.8 with the remaining GWAS SNPs.

Supplementary Material

1
2
3
4
5
6

  • Analyses of variations in histone marks reveal histone QTLs in regulatory elements

  • Physically interacting loci show genetically coordinated chromatin levels

  • Regulatory elements sharing hQTLs are enriched in topologically associated domains

  • hQTLs are enriched for GWAS SNPs and enable identification of putative target genes

ACKNOWLEDGMENTS

Our work was supported by grants from the Vera Moulton Wall Center for Pulmonary Vascular Disease (F.G., M.K.), Swiss National Foundation (J.B.Z.), NIH Medical Scientist Training Program Training Grant T32GM007205 (M.K.), Gabilan Stanford Graduate Fellowship (O.U.), NIH/NHGRI T32 HG000044 and the Genentech Graduate Fellowship (D.V.S), NIH Genetics and Developmental Biology Training Program T32 GM007790 (A.R.M), Stanford Biomedical Informatics Training Grant from the National Library of Medicine (LM-07033) (P.G.), R.S. is a Damon Runyon Fellow supported by the Damon Runyon Cancer Research Foundation (DRG-2187-14), D.H.P. is a Damon Runyon fellow supported by the Damon Runyon Cancer Research Foundation (DRG-2122-12), EC FP7 project “RADIANT” (A.P., W.H.), NIH and from the European Research Council (L.M.S), Alfred Sloan Foundation and NIH R01ES025009 and NIH U41-HG007000-02S1 (subcontract) (A.K.), NIH and Genetics Department, Stanford University (M.S.)

C.D.B. is a founder of Identify Genomics and a science advisory board member for AncestryDNA, InVitae, Personalis and Identify Genomics; L.M.S. is a founder and member of the science advisory board for Sophia Genetics; M.S. is a founder and member of the science advisory board of Personalis and a science advisory board member of Genapsys and AxioMX.

We would like to thank Prof. Stephen Montgomery and members of the Snyder lab for critical comments on the manuscript. We would also like to thank Gerald Quon and Abhishek Sarkar for sharing GWAS summary statistics and SCGPM and SCG3 administrators for compute support and supplementary website hosting.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Genetic variation in regulatory elements can effect coordinate changes in chromatin state and gene expression at both local and distal sites reflecting associations in a three-dimensional context. Integrating information from expression, chromatin modification and chromosome contact analyses provides a framework for assessing disease-associated mutations.

ACCESSION NUMBERS

The Gene Expression Omnibus (GEO) accession number for the data sets reported in this paper is GSE62742. The data and analysis files are at http://chromovar3d.stanford.edu/ and http://www.zaugg.embl.de/data-and-tools/distal-chromatin-qtls/, code at https://github.com/kundajelab/chromovar3d

AUTHOR CONTRIBUTION

Conceptualization, F.G., J.B.Z., M.K., O.U.; Methodology, F.G., J.B.Z., M.K., O.U., D.S., A.R.M., P.G., R.S., D.H.P., A.P., N.H., G.E.; Investigation, F.G., M.K., D.S.; Formal Analysis, J.B.Z., F.G., O.U., M.K., A.R.M., P.G., R.S., D.H.P.; Writing – Original Draft, F.G., J.B.Z., M.K., O.U., A.R.M., P.G., R.S., A.K., M.S. ; Writing – Review & Editing, F.G., J.B.Z., M.K., O.U., A.R.M., P.G., R.S., D.H.P., D.S., A.K., M.S; Funding Acquisition, M.S.; Resources, M.S., A.K.; Supervision, M.S., A.K., L.M.S., C.D.B., J.K.P., W.H.

REFERENCES

  1. Bourgon R, Gentleman R, Huber W. Independent filtering increases detection power for high-throughput experiments. Proc Natl Acad Sci U S A. 2010;107:9546–9551. doi: 10.1073/pnas.0914005107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Degner JF, Pai AA, Pique-Regi R, Veyrieras JB, Gaffney DJ, Pickrell JK, De Leon S, Michelini K, Lewellen N, Crawford GE, et al. DNase I sensitivity QTLs are a major determinant of human expression variation. Nature. 2012;482:390–394. doi: 10.1038/nature10808. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Dixon JR, Selvaraj S, Yue F, Kim A, Li Y, Shen Y, Hu M, Liu JS, Ren B. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature. 2012;485:376–380. doi: 10.1038/nature11082. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Dunham I, Kundaje A, Aldred SF, Collins PJ, Davis CA, Doyle F, Epstein CB, Frietze S, Harrow J, Kaul R, et al. An integrated encyclopedia of DNA elements in the human genome. Nature. 2012;489:57–74. doi: 10.1038/nature11247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ernst J, Kheradpour P, Mikkelsen TS, Shoresh N, Ward LD, Epstein CB, Zhang X, Wang L, Issner R, Coyne M, et al. Mapping and analysis of chromatin state dynamics in nine human cell types. Nature. 2011;473:43–49. doi: 10.1038/nature09906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Franke A, McGovern DP, Barrett JC, Wang K, Radford-Smith GL, Ahmad T, Lees CW, Balschun T, Lee J, Roberts R, et al. Genome-wide meta-analysis increases to 71 the number of confirmed Crohn’s disease susceptibility loci. Nat Genet. 2010;42:1118–1125. doi: 10.1038/ng.717. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Fullwood MJ, Liu MH, Pan YF, Liu J, Xu H, Mohamed YB, Orlov YL, Velkov S, Ho A, Mei PH, et al. An oestrogen-receptor-alpha-bound human chromatin interactome. Nature. 2009;462:58–64. doi: 10.1038/nature08497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Gjoneska E, Pfenning AR, Mathys H, Quon G, Kundaje A, Tsai LH, Kellis M. Conserved epigenomic signals in mice and humans reveal immune basis of Alzheimer’s disease. Nature. 2015;518:365–369. doi: 10.1038/nature14252. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Harismendy O, Notani D, Song X, Rahim NG, Tanasa B, Heintzman N, Ren B, Fu XD, Topol EJ, Rosenfeld MG, et al. 9p21 DNA variants associated with coronary artery disease impair interferon-gamma signalling response. Nature. 2011;470:264–268. doi: 10.1038/nature09753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Harrow J, Frankish A, Gonzalez JM, Tapanari E, Diekhans M, Kokocinski F, Aken BL, Barrell D, Zadissa A, Searle S, et al. GENCODE: the reference human genome annotation for The ENCODE Project. Genome Res. 2012;22:1760–1774. doi: 10.1101/gr.135350.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hindorff LA, Sethupathy P, Junkins HA, Ramos EM, Mehta JP, Collins FS, Manolio TA. Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proc Natl Acad Sci U S A. 2009;106:9362–9367. doi: 10.1073/pnas.0903103106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Jin F, Li Y, Dixon JR, Selvaraj S, Ye Z, Lee AY, Yen CA, Schmitt AD, Espinoza CA, Ren B. A high-resolution map of the three-dimensional chromatin interactome in human cells. Nature. 2013;503:290–294. doi: 10.1038/nature12644. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Karczewski KJ, Dudley JT, Kukurba KR, Chen R, Butte AJ, Montgomery SB, Snyder M. Systematic functional regulatory assessment of disease-associated variants. Proc Natl Acad Sci U S A. 2013;110:9607–9612. doi: 10.1073/pnas.1219099110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kasowski M, Kyriazopoulou-Panagiotopoulou S, Grubert F, Zaugg JB, Kundaje A, Liu Y, Boyle AP, Zhang QC, Zakharia F, Spacek DV, et al. Extensive variation in chromatin states across humans. Science. 2013;342:750–752. doi: 10.1126/science.1242510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kheradpour P, Kellis M. Systematic discovery and characterization of regulatory motifs in ENCODE TF binding experiments. Nucleic acids research. 2014;42:2976–2987. doi: 10.1093/nar/gkt1249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Kleinjan DA, van Heyningen V. Long-range control of gene expression: emerging mechanisms and disruption in disease. American journal of human genetics. 2005;76:8–32. doi: 10.1086/426833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Koren A, Polak P, Nemesh J, Michaelson JJ, Sebat J, Sunyaev SR, McCarroll SA. Differential relationship of DNA replication timing to different forms of human mutation and variation. American journal of human genetics. 2012;91:1033–1040. doi: 10.1016/j.ajhg.2012.10.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Kundaje A, Kyriazopoulou-Panagiotopoulou S, Libbrecht M, Smith CL, Raha D, Winters EE, Johnson SM, Snyder M, Batzoglou S, Sidow A. Ubiquitous heterogeneity and asymmetry of the chromatin environment at regulatory elements. Genome Res. 2012;22:1735–1747. doi: 10.1101/gr.136366.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Lambert JC, Ibrahim-Verbaas CA, Harold D, Naj AC, Sims R, Bellenguez C, DeStafano AL, Bis JC, Beecham GW, Grenier-Boley B, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet. 2013;45:1452–1458. doi: 10.1038/ng.2802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Lappalainen T, Sammeth M, Friedlander MR, t Hoen PA, Monlong J, Rivas MA, Gonzalez-Porta M, Kurbatova N, Griebel T, Ferreira PG, et al. Transcriptome and genome sequencing uncovers functional variation in humans. Nature. 2013;501:506–511. doi: 10.1038/nature12531. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Li G, Ruan X, Auerbach RK, Sandhu KS, Zheng M, Wang P, Poh HM, Goh Y, Lim J, Zhang J, et al. Extensive promoter-centered chromatin interactions provide a topological basis for transcription regulation. Cell. 2012;148:84–98. doi: 10.1016/j.cell.2011.12.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Lieberman-Aiden E, van Berkum NL, Williams L, Imakaev M, Ragoczy T, Telling A, Amit I, Lajoie BR, Sabo PJ, Dorschner MO, et al. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science. 2009;326:289–293. doi: 10.1126/science.1181369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Maurano MT, Humbert R, Rynes E, Thurman RE, Haugen E, Wang H, Reynolds AP, Sandstrom R, Qu H, Brody J, et al. Systematic localization of common disease-associated variation in regulatory DNA. Science. 2012;337:1190–1195. doi: 10.1126/science.1222794. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Nicolae DL, Gamazon E, Zhang W, Duan S, Dolan ME, Cox NJ. Trait-associated SNPs are more likely to be eQTLs: annotation to enhance discovery from GWAS. PLoS Genet. 2010;6:e1000888. doi: 10.1371/journal.pgen.1000888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nobrega MA, Ovcharenko I, Afzal V, Rubin EM. Scanning human gene deserts for long-range enhancers. Science. 2003;302:413. doi: 10.1126/science.1088328. [DOI] [PubMed] [Google Scholar]
  26. Patro R, Mount SM, Kingsford C. Sailfish enables alignment-free isoform quantification from RNA-seq reads using lightweight algorithms. Nat Biotechnol. 2014;32:462–464. doi: 10.1038/nbt.2862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Phanstiel DH, Boyle AP, Heidari N, Snyder MP. Mango: A bias correcting ChIA-PET analysis pipeline. Bioinformatics. 2015 doi: 10.1093/bioinformatics/btv336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Pomerantz MM, Ahmadiyeh N, Jia L, Herman P, Verzi MP, Doddapaneni H, Beckwith CA, Chan JA, Hills A, Davis M, et al. The 8q24 cancer risk variant rs6983267 shows long-range interaction with MYC in colorectal cancer. Nat Genet. 2009;41:882–884. doi: 10.1038/ng.403. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rao SS, Huntley MH, Durand NC, Stamenova EK, Bochkov ID, Robinson JT, Sanborn AL, Machol I, Omer AD, Lander ES, et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell. 2014;159:1665–1680. doi: 10.1016/j.cell.2014.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Rivas MA, Beaudoin M, Gardet A, Stevens C, Sharma Y, Zhang CK, Boucher G, Ripke S, Ellinghaus D, Burtt N, et al. Deep resequencing of GWAS loci identifies independent rare variants associated with inflammatory bowel disease. Nat Genet. 2011;43:1066–1073. doi: 10.1038/ng.952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sanyal A, Lajoie BR, Jain G, Dekker J. The long-range interaction landscape of gene promoters. Nature. 2012;489:109–113. doi: 10.1038/nature11279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Schaub MA, Boyle AP, Kundaje A, Batzoglou S, Snyder M. Linking disease associations with regulatory information in the human genome. Genome Res. 2012;22:1748–1759. doi: 10.1101/gr.136127.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Smemo S, Tena JJ, Kim KH, Gamazon ER, Sakabe NJ, Gomez-Marin C, Aneas I, Credidio FL, Sobreira DR, Wasserman NF, et al. Obesity-associated variants within FTO form long-range functional connections with IRX3. Nature. 2014;507:371–375. doi: 10.1038/nature13138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Stegle O, Parts L, Durbin R, Winn J. A Bayesian framework to account for complex non-genetic factors in gene expression levels greatly increases power in eQTL studies. PLoS computational biology. 2010;6:e1000770. doi: 10.1371/journal.pcbi.1000770. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Van Limbergen J, Wilson DC, Satsangi J. The genetics of Crohn’s disease. Annual review of genomics and human genetics. 2009;10:89–116. doi: 10.1146/annurev-genom-082908-150013. [DOI] [PubMed] [Google Scholar]
  36. Visel A, Rubin EM, Pennacchio LA. Genomic views of distant-acting enhancers. Nature. 2009;461:199–205. doi: 10.1038/nature08451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, Bernstein BE, Nusbaum C, Myers RM, Brown M, Li W, et al. Model-based analysis of ChIP-Seq (MACS) Genome Biol. 2008;9:R137. doi: 10.1186/gb-2008-9-9-r137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Zuk O, Schaffner SF, Samocha K, Do R, Hechter E, Kathiresan S, Daly MJ, Neale BM, Sunyaev SR, Lander ES. Searching for missing heritability: designing rare variant association studies. Proc Natl Acad Sci U S A. 2014;111:E455–E464. doi: 10.1073/pnas.1322563111. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS717450-supplement-1.pdf (234.2KB, pdf)
2
NIHMS717450-supplement-2.docx (106.4KB, docx)
3
NIHMS717450-supplement-3.xls (69KB, xls)
4
NIHMS717450-supplement-4.xlsx (30.2MB, xlsx)
5
NIHMS717450-supplement-5.xlsx (6.8MB, xlsx)
6
NIHMS717450-supplement-6.xlsx (1.8MB, xlsx)

RESOURCES