Introduction

CHiCAGO is a method for detecting statistically significant interaction events in Capture HiC data. This vignette will walk you through a typical CHiCAGO analysis.

A typical Chicago job for two biological replicates of CHi-C data takes 2-3 h wall-clock time (including sample pre-processing from bam files) and uses 50G RAM.

NOTE A wrapper to perform this type of analysis, called runChicago.R, is provided as part of chicagoTools, which is available from our Bitbucket repository. Refer to the chicagoTools README for more information.

The statistical foundations of CHiCAGO are presented in a separate paper that is currently available as a preprint (Jonathan Cairns*, Paula Freire-Pritchett* et al. 2015). Briefly, CHiCAGO uses a convolution background model accounting for both ‘Brownian collisions’ between fragments (distance-dependent) and ‘technical noise’. It borrows information across interactions (with appropriate normalisation) to estimate these background components separately on different subsets of data. CHiCAGO then uses a p-value weighting procedure based on the expected true positive rates at different distance ranges (estimated from data), with scores representing soft-thresholded -log weighted p-values. The score threshold of 5 is a suggested stringent score threshold for calling significant interactions.

WARNING The data set used in this tutorial comes from the package PCHiCdata. This package contains small parts (two chromosomes each) of published Promoter Capture HiC data sets in mouse ESCs (Schoenfelder et al. 2015) and GM12878 cells, derived from human LCLs (Mifsud et al. 2015) - note that both papers used a different interaction-calling algorithm and we are only reusing raw data from them. Do not use any of these sample input data for purposes other than training.

In this vignette, we use the GM12878 data (Mifsud et al. 2015):

library(Chicago)
library(PCHiCdata)

Input files required

Before you start, you will need:

  1. Five restriction map information files (“design files”):

We recommend that you put all five of these files into the same directory (that we refer to as designDir). An example of a valid design folder, for a two-chromosome sample of the GM12878 data used in this vignette, is provided in the PCHiCdata package, as follows.

dataPath <- system.file("extdata", package="PCHiCdata")
testDesignDir <- file.path(dataPath, "hg19TestDesign")
dir(testDesignDir)
## [1] "h19_chr20and21.baitmap" "h19_chr20and21.nbpb"   
## [3] "h19_chr20and21.npb"     "h19_chr20and21.poe"    
## [5] "h19_chr20and21.rmap"
NOTE Though we talk about “restriction fragments” throughout, any non-overlapping regions can be defined in .rmap (with a subset of these defined as baits). For example, if one wanted to increase power at the cost of resolution, it is possible to use bins of restriction fragments either throughout, or for some selected regions. However, in the binned fragment framework, we advise setting removeAdjacent to FALSE - see ?setExperiment for details on how to do this.
  1. You will also need input data files. These should be in CHiCAGO input format, .chinput. .chinput files can be obtained from aligned Capture HiC BAM files by running bam2chicago.sh, available as part of chicagoTools. (To obtain BAM files from raw fastq files, run them through a HiC alignment & QC pipeline such as HiCUP.

Example .chinput files are provided in the PCHiCdata package, as follows:

testDataPath <- file.path(dataPath, "GMchinputFiles")
dir(testDataPath)
## [1] "GM_rep1.chinput" "GM_rep2.chinput" "GM_rep3.chinput"
files <- c(
    file.path(testDataPath, "GM_rep1.chinput"),
    file.path(testDataPath, "GM_rep2.chinput"),
    file.path(testDataPath, "GM_rep3.chinput")
  )

OPTIONAL: The data set in this vignette requires some additional custom settings, to ensure that the vignette compiles in a reasonable time:

settingsFile <- file.path(system.file("extdata", package="PCHiCdata"),
                          "sGM12878Settings", "sGM12878.settingsFile")

Omit this step unless you know which settings you wish to alter. If you are using non-human samples, or a very unusual cell type, one set of options that you might want to change is the weighting parameters: see Using different weights.

Example workflow

We run CHiCAGO on the test data as follows. First, we create a blank chicagoData object, and we tell it where the design files are. For this example, we also provide the optional settings file - in your analysis, you can omit the settingsFile argument.

library(Chicago)

cd <- setExperiment(designDir = testDesignDir, settingsFile = settingsFile)

The properties of chicagoData objects are discussed more in The chicagoData object.

Next, we read in the input data files:

cd <- readAndMerge(files=files, cd=cd)

Finally, we run the pipeline with chicagoPipeline():

cd <- chicagoPipeline(cd)

Output plots

chicagoPipeline() produces a number of plots. You can save these to disk by setting the outprefix argument in chicagoPipeline().

The plots are as follows (an explanation follows):

Interpreting the plots

In this section, we describe the expected properties of the diagnostic plots.

Note that the diagnostic plots above are computed on the fly using only a small fraction of the full GM12878 dataset. In real-world, genome-wide datasets, more fragment pools will be visible and thus the trends described below should be more pronounced.

  1. Brownian other end factors: The adjustment made to the mean Brownian read count, estimated in the pools of other ends. (“tlb” refers to the number of trans-chromosomal reads that the other end accumulates in total. “B2B” stands for a “bait-to-bait” interactions).
  1. Technical noise estimates: The mean number of technical noise reads expected for other ends and baits, respectively, per pools of fragments. These pools, displayed on the x axis, again refer to the number of trans-chromosomal reads that the fragments accumulate.
  1. Distance function: The mean number of Brownian reads expected for an average bait, as a function of distance, plotted on a log-log scale.

Output files

Two main output methods are provided. Here, we discuss the first: exporting to disk. However, it is also possible to export to a GenomeInteractions object, as described in Further downstream analysis.

You can export the results to disk, using exportResults(). (If you use runChicago.R, the files appear in ./<results-folder>/data). By default, the function outputs three different output file formats:

outputDirectory <- tempdir()
exportResults(cd, file.path(outputDirectory, "vignetteOutput"))
## Reading the restriction map file...
## Reading the bait map file...
## Preparing the output table...
## Writing out for seqMonk...
## Writing out interBed...
## Preprocessing for WashU outputs...
## Writing out text file for WashU browser upload...

Each called interaction is assigned a score that represents how strong CHiCAGO believes the interaction is (formally, it is based on -log(adjusted P-value)). Thus, a larger score represents a stronger interaction. In each case, the score threshold of 5 is applied.

Summary of output files:

ibed format (ends with …ibed):

##   bait_chr bait_start bait_end                  bait_name otherEnd_chr
## 1       20     119103   138049                    DEFB126           20
## 2       20     161620   170741                    DEFB128           20
## 3       20     233983   239479                    DEFB132           20
## 4       20     268639   284501 AL034548.1;C20orf96;ZCCHC3           20
## 5       20     268639   284501 AL034548.1;C20orf96;ZCCHC3           20
## 6       20     305699   309123         RP5-1103G7.4;SOX12           20
##   otherEnd_start otherEnd_end otherEnd_name N_reads score
## 1         523682       536237       CSNK2A1       6  6.00
## 2          73978        76092             .      16  5.17
## 3         206075       209203       DEFB129      33  5.35
## 4         293143       304037             .      34  7.26
## 5         304038       305698             .      34  8.77
## 6         293143       304037             .      37  5.49

seqmonk format (ends with …seqmonk.txt):

##   V1     V2     V3      V4 V5   V6
## 1 20 119103 138049 DEFB126  6 6.00
## 2 20 523682 536237 CSNK2A1  6 6.00
## 3 20 161620 170741 DEFB128 16 5.17
## 4 20  73978  76092       . 16 5.17
## 5 20 233983 239479 DEFB132 33 5.35
## 6 20 206075 209203 DEFB129 33 5.35

washU_text format (ends with …washU_text.txt):

##                    V1                  V2   V3
## 1 chr20,119103,138049 chr20,523682,536237 6.00
## 2 chr20,161620,170741   chr20,73978,76092 5.17
## 3 chr20,233983,239479 chr20,206075,209203 5.35
## 4 chr20,268639,284501 chr20,293143,304037 7.26
## 5 chr20,268639,284501 chr20,304038,305698 8.77
## 6 chr20,305699,309123 chr20,293143,304037 5.49

For bait-to-bait interactions, the interaction can be tested either way round (i.e. either fragment can be considered the “bait”). In most output formats, both of these tests are preserved. The exception is washU output, where these scores are consolidated by taking the maximum.

NOTE When comparing interactions detected between multiple replicates, the degree of overlap may appear to be lower than expected. This is due to the undersampled nature of most CHi-C datasets. Sampling error can drive down the sensitivity, particularly for interactions that span large distances and have low read counts. As such, low overlap is not necessarily an indication of a high false discovery rate.

Undersampling needs to be taken into consideration when interpreting CHiCAGO results. In particular, we recommend performing comparisons at the score-level rather than at the level of thresholded interaction calls. Potentially, differential analysis algorithms for sequencing data such as DESeq2 (Love, Huber, and Anders 2014) may also be used to formally compare the enrichment at CHiCAGO-detected interactions between conditions at the count level, although power will generally be a limiting factor.

Formal methods such as sdef (Blangiardo, Cassese, and Richardson 2010) may provide a more balanced view of the consistency between replicates. Alternatively, additional filtering based on the mean number of reads per detected interaction (e.g. removing calls with N<10 reads) will reduce the impact of undersampling on the observed overlap, but at the cost of decreasing the power to detect longer-range interactions.

Visualising interactions

The plotBaits() function can be used to plot the raw read counts versus linear distance from bait for either specific or random baits, labelling significant interactions in a different colour. By default, 16 random baits are plotted, with interactions within 1 Mb from bait passing the threshold of 5 shown in red and those passing the more lenient threshold of 3 shown in blue.

plottedBaitIDs <- plotBaits(cd, n=6)

Peak enrichment for features

peakEnrichment4Features() tests the hypothesis that other ends in the CHiCAGO output are enriched for genomic features of interest - for example, histone marks associated with enhancers. We find out how many overlaps are expected under the null hypothesis (i.e. that there is no enrichment) by shuffling the other ends around in the genome, while preserving the overall distribution of distances over which interactions span.

You will need additional files to perform this analysis - namely, a .bed file for each feature. We include ChIP-seq data from the ENCODE consortium (The ENCODE Project Consortium 2012), also restricted to chr20 and chr21. (Data accession numbers: Bernstein lab GSM733752, GSM733772, GSM733708, GSM733664, GSM733771, GSM733758)

First, we find the folder that contains the features, and construct a list of the features to use:

featuresFolder <- file.path(dataPath, "GMfeatures")
dir(featuresFolder)
## [1] "featuresGM.txt"                                                 
## [2] "spp.wgEncodeBroadHistoneGm12878CtcfStdAln_chr20and21.narrowPeak"
## [3] "wgEncodeBroadHistoneGm12878H3k27acStdAln_chr20and21.narrowPeak" 
## [4] "wgEncodeBroadHistoneGm12878H3k27me3StdAln_chr20and21.narrowPeak"
## [5] "wgEncodeBroadHistoneGm12878H3k4me1StdAln_chr20and21.narrowPeak" 
## [6] "wgEncodeBroadHistoneGm12878H3k4me3StdAln_chr20and21.narrowPeak" 
## [7] "wgEncodeBroadHistoneGm12878H3k9me3StdAln_chr20and21.narrowPeak"
featuresFile <- file.path(featuresFolder, "featuresGM.txt")
featuresTable <- read.delim(featuresFile, header=FALSE, as.is=TRUE)
featuresList <- as.list(featuresTable$V2)
names(featuresList) <- featuresTable$V1
featuresList
## $CTCF
## [1] "spp.wgEncodeBroadHistoneGm12878CtcfStdAln_chr20and21.narrowPeak"
## 
## $H3K4me1
## [1] "wgEncodeBroadHistoneGm12878H3k4me1StdAln_chr20and21.narrowPeak"
## 
## $H3K4me3
## [1] "wgEncodeBroadHistoneGm12878H3k4me3StdAln_chr20and21.narrowPeak"
## 
## $H3k27ac
## [1] "wgEncodeBroadHistoneGm12878H3k27acStdAln_chr20and21.narrowPeak"
## 
## $H3K27me3
## [1] "wgEncodeBroadHistoneGm12878H3k27me3StdAln_chr20and21.narrowPeak"
## 
## $H3K9me3
## [1] "wgEncodeBroadHistoneGm12878H3k9me3StdAln_chr20and21.narrowPeak"

Next, we feed this information into the peakEnrichment4Features() function.

As part of the analysis, peakEnrichment4Features() takes a distance range (by default, the full distance range over which interactions are observed), and divides it into some number of bins. We must select the number of bins; here, we choose that number to ensure that the bin size is approximately 10kb. If the defaults are changed, a different number of bins is more appropriate. See ?peakEnrichment4Features for more information.

no_bins <- ceiling(max(abs(intData(cd)$distSign), na.rm = TRUE)/1e4)

enrichmentResults <- peakEnrichment4Features(cd, folder=featuresFolder,
              list_frag=featuresList, no_bins=no_bins, sample_number=100)

Note the plot produced by this function. For each feature type, the yellow bar represents the number of features that overlap with interaction other ends. The blue bar represents what would be expected by chance, with a 95% confidence interval for the mean number of overlaps plotted. If the yellow bar lies outside of this interval, we reject the null hypothesis, thus concluding that there is enrichment/depletion of that feature.

The information displayed in the plot is also returned in tabular form (OL = Overlap, SI = Significant Interactions, SD = Standard Deviation, CI = Confidence Interval):

enrichmentResults
##          OLwithSI MeanOLwithSamples SDOLwithSample   LowerCI  HigherCI
## CTCF          345            122.23      10.661793 101.33288 143.12712
## H3K4me1       685            270.44      14.104459 242.79526 298.08474
## H3K4me3       373            135.30       9.472448 116.73400 153.86600
## H3k27ac       467            164.42      11.373529 142.12788 186.71212
## H3K27me3       71             71.62       7.485717  56.94799  86.29201
## H3K9me3       226            120.21       9.753834 101.09249 139.32751

Further downstream analysis

We can perform further downstream analysis in R or Bioconductor, using functionality from the GenomicInteractions package. First, we export the significant interactions into a GenomicInteractions object:

library(GenomicInteractions)
## Warning: replacing previous import by 'ggplot2::Position' when loading
## 'GenomicInteractions'
library(GenomicRanges)
gi <- exportToGI(cd)

From here, we can pass the CHiCAGO results through to other Bioconductor functionality. In the following example, we find out which other ends overlap with the H3K4me1 enhancer mark, using ENCODE data. We use AnnotationHub to fetch a relevant enhancer mark track from the ENCODE project:

library(AnnotationHub)
ah <- AnnotationHub()
hs <- query(ah, c("GRanges", "EncodeDCC", "Homo sapiens", "H3k4me1"))
enhancerTrack <- hs[["AH23254"]]

Next, we use the anchorTwo() function to extract the other end locations from the GenomicInteractions object (anchorOne() would give us the bait locations instead). Note that in this particular instance, the seqlevels() also need to be changed before performing the comparison, adding “chr” to make them match those of the annotation.

otherEnds <- anchorTwo(gi)
otherEnds <- renameSeqlevels(otherEnds, c("chr20","chr21"))

Finally, we look at which other ends overlap the enhancer marks:

findOverlaps(otherEnds, enhancerTrack)
## Hits object with 4556 hits and 0 metadata columns:
##          queryHits subjectHits
##          <integer>   <integer>
##      [1]         1       40732
##      [2]         1       40735
##      [3]         8       40717
##      [4]         8       40718
##      [5]         9       40721
##      ...       ...         ...
##   [4552]      4160       17630
##   [4553]      4160       17631
##   [4554]      4161       17676
##   [4555]      4161       17677
##   [4556]      4164       17665
##   -------
##   queryLength: 4165
##   subjectLength: 109612

Further note that the annotation’s genome version should match that of the promoter capture data, namely hg19:

hs["AH23254"]$genome
## [1] "hg19"

The chicagoData object

In the above workflows, cd is a chicagoData object. It contains three elements:

A closer look at intData(cd):

head(intData(cd), 2)
##    baitID otherEndID distbin       s_j otherEndLen distSign isBait2bait
## 1: 403463     403833      NA 0.2368791        2579  1652804       FALSE
## 2: 403463     403843      NA 0.2368791        6302  1690808       FALSE
##    N.1 N.2 N.3 N refBinMean       s_i NNb NNboe    tlb      tblb
## 1:   0   1   0 1         NA 0.9494934   4     4  [0,6] [  2, 46)
## 2:   0   1   0 1         NA 1.0021336   4     4 (6,13] [  2, 46)
##          Tmean      Bmean     log.p    log.w     log.q score
## 1: 0.001485762 0.08970408 -2.443292 1.406451 -3.849743     0
## 2: 0.004021026 0.09243682 -2.389726 1.350609 -3.740334     0

Columns:

WARNING: Many functions in CHiCAGO update intData(cd) by reference, which means that intData(cd) can change even when you do not explicitly assign to it. To avoid this behaviour, copy the chicagoData object first:

newCd = copyCD(cd)

Using different weights

CHiCAGO uses a p-value weighting procedure to upweight proximal interactions and downweight distal interactions. This procedure has four tuning parameters.

The default values of these tuning parameters were calibrated on calls from seven human Macrophage data sets. Provided that your cell type is not too dissimilar to these calibration data, it should be fine to leave the parameters at their default settings. However, if your data set is from a different species or an unusual cell type, you may wish to recalibrate these parameters using data from cell types similar to yours. You can do this with the fitDistCurve.R script in chicagoTools, which we demonstrate in this section.

First, run all of the samples through chicagoPipeline(), saving each chicagoData object in individual .rds files (see saveRDS()). Alternatively, if you are using the runChicago.R wrapper, .rds files should be generated automatically.

Second, run the fitDistCurve.R script. As an example, if we had three biological replicates of mESC cells, we might run the following script at the Unix command prompt:

Rscript fitDistCurve.R mESC --inputs mESC1.rds,mESC2.rds,mESC3.rds 

This script produces the file mESC.settingsFile, which you can read in to modifySettings() as usual - see the Input files required section.

Additionally, the script produces a plot, mESC_mediancurveFit.pdf, that can be used to diagnose unreliable estimates. By default, five coloured lines are shown, each representing a parameter estimate from a subset of the data. An unreliable fit is typically diagnosed when the coloured lines are highly dissimilar to each other, and thus the black median line is not representative of them. (Some dissimilarity may be OK, since the median confers robustness.)

For the user’s convenience, a set of precomputed weights are also provided in the package:

weightsPath <- file.path(system.file("extdata", package="Chicago"),
                         "weights")
dir(weightsPath)
## [1] "GM12878-2reps.settings"         "humanMacrophage-7reps.settings"
## [3] "mESC-2reps.settings"

For example, to use the GM12878 weights, supply the appropriate settings file to setExperiment() as per the following:

weightSettings <- file.path(weightsPath, "GM12878-2reps.settings")
cd <- setExperiment(designDir = testDesignDir, settingsFile = weightSettings)

Session info

sessionInfo()
## R version 3.2.2 (2015-08-14)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Ubuntu 14.04.4 LTS
## 
## locale:
##  [1] LC_CTYPE=en_GB.UTF-8       LC_NUMERIC=C              
##  [3] LC_TIME=en_GB.UTF-8        LC_COLLATE=C              
##  [5] LC_MONETARY=en_GB.UTF-8    LC_MESSAGES=en_GB.UTF-8   
##  [7] LC_PAPER=en_GB.UTF-8       LC_NAME=C                 
##  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
## [11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C       
## 
## attached base packages:
## [1] stats4    parallel  stats     graphics  grDevices utils     datasets 
## [8] methods   base     
## 
## other attached packages:
##  [1] rtracklayer_1.28.10       AnnotationHub_2.0.4      
##  [3] GenomicRanges_1.20.8      GenomeInfoDb_1.4.3       
##  [5] IRanges_2.2.9             S4Vectors_0.6.6          
##  [7] BiocGenerics_0.14.0       GenomicInteractions_1.2.3
##  [9] PCHiCdata_0.99.3          Chicago_0.99.6           
## [11] data.table_1.9.6          BiocStyle_1.6.0          
## 
## loaded via a namespace (and not attached):
##  [1] Rcpp_0.12.4                  lattice_0.20-33             
##  [3] Delaporte_2.2-3              Rsamtools_1.20.5            
##  [5] Biostrings_2.36.4            assertthat_0.1              
##  [7] digest_0.6.9                 mime_0.4                    
##  [9] R6_2.1.2                     plyr_1.8.3                  
## [11] chron_2.3-47                 futile.options_1.0.0        
## [13] acepack_1.3-3.3              RSQLite_1.0.0               
## [15] evaluate_0.8.3               BiocInstaller_1.18.5        
## [17] httr_1.1.0                   ggplot2_2.1.0               
## [19] zlibbioc_1.14.0              curl_0.9.6                  
## [21] rpart_4.1-10                 rmarkdown_0.9.5             
## [23] splines_3.2.2                BiocParallel_1.2.22         
## [25] stringr_1.0.0                foreign_0.8-66              
## [27] igraph_1.0.1                 RCurl_1.95-4.8              
## [29] munsell_0.4.3                shiny_0.13.2                
## [31] httpuv_1.3.3                 htmltools_0.3.5             
## [33] nnet_7.3-12                  gridExtra_2.2.1             
## [35] interactiveDisplayBase_1.6.1 Hmisc_3.17-2                
## [37] matrixStats_0.50.1           XML_3.98-1.4                
## [39] dplyr_0.4.3                  GenomicAlignments_1.4.2     
## [41] MASS_7.3-45                  bitops_1.0-6                
## [43] grid_3.2.2                   xtable_1.8-2                
## [45] gtable_0.2.0                 DBI_0.3.1                   
## [47] magrittr_1.5                 formatR_1.3                 
## [49] scales_0.4.0                 stringi_1.0-1               
## [51] XVector_0.8.0                latticeExtra_0.6-28         
## [53] futile.logger_1.4.1          Formula_1.2-1               
## [55] lambda.r_1.1.7               RColorBrewer_1.1-2          
## [57] tools_3.2.2                  Biobase_2.28.0              
## [59] plotrix_3.6-1                survival_2.38-3             
## [61] yaml_2.1.13                  AnnotationDbi_1.30.1        
## [63] colorspace_1.2-6             cluster_2.0.3               
## [65] knitr_1.12.3

References

Blangiardo, Marta, Alberto Cassese, and Sylvia Richardson. 2010. “sdef: an R Package to Synthesize Lists of Significant Features in Related Experiments.” BMC Bioinformatics 11 (1): 270.

Jonathan Cairns*, Paula Freire-Pritchett*, Steven W. Wingett, Andrew Dimond, Vincent Plagnol, Daniel Zerbino, Stefan Schoenfelder, Biola-Maria Javierre, Cameron Osborne, Peter Fraser, and Mikhail Spivakov. 2015. “CHiCAGO: Robust Detection of DNA Looping Interactions in Capture Hi-C Data.” bioRxiv. doi:10.1101/028068.

Love, Michael I, Wolfgang Huber, and Simon Anders. 2014. “Moderated Estimation of Fold Change and Dispersion for RNA-Seq Data with DESeq2.” Genome Biology 15 (12): 550. doi:10.1186/s13059-014-0550-8. http://dx.doi.org/10.1186/s13059-014-0550-8.

Mifsud, Borbala, Filipe Tavares-Cadete, Alice N Young, Robert Sugar, Stefan Schoenfelder, Lauren Ferreira, Steven W Wingett, et al. 2015. “Mapping Long-Range Promoter Contacts in Human Cells with High-Resolution Capture Hi-C.” Nature Genetics 47 (6): 598–606.

Schoenfelder, Stefan, Mayra Furlan-Magaril, Borbala Mifsud, Filipe Tavares-Cadete, Robert Sugar, Biola-Maria Javierre, Takashi Nagano, et al. 2015. “The Pluripotent Regulatory Circuitry Connecting Promoters to Their Long-Range Interacting Elements.” Genome Research 25 (4): 582–597.

The ENCODE Project Consortium. 2012. “An Integrated Encyclopedia of DNA Elements in the Human Genome.” Nature 489 (7414) (September): 57–74.