knitr::opts_chunk$set( warning=FALSE, message=FALSE )

1 Biscuiteer

biscuiteer is package to process output from biscuit into bsseq objects. It includes a number of features, such as VCF header parsing, shrunken M-value calculations (which can be used for compartment inference), and age inference. However, the task of locus- and region-level differential methylation inference is delegated to other packages (such as dmrseq).

2 Quick Start

2.1 Installing

From Bioconductor,

if (!requireNamespace("BiocManager", quietly=TRUE))
    install.packages("BiocManager")
BiocManager::install("biscuiteer")

A development version is available on GitHub and can be installed via:

if (!requireNamespace("BiocManager", quietly=TRUE))
    install.packages("BiocManager")
BiocManager::install("trichelab/biscuiteerData")
BiocManager::install("trichelab/biscuiteer")

2.2 Loading Methylation Data

biscuiteer can load either headered or header-free BED files produced from biscuit vcf2bed or biscuit mergecg. In either case, a VCF file is needed when loading biscuit output. For practical purposes, only the VCF header is for biscuiteer. However, it is encouraged that the user keep the entire VCF, as biscuit can be used to call SNVs and allows for structural variant detection in a similar manner to typical whole-genome sequencing tools. Furthermore, biscuit records the version of the software and the calling arguments used during processing the output VCF, which allows for better reproducibility.

NOTE: Both the input BED and VCF files must be tabix’ed before being input to biscuiteer. This can be done by running bgzip biscuit_output.xxx followed by tabix -p xxx biscuit_output.xxx.gz, where xxx is either bed or vcf.

Data can be loaded using the readBiscuit function in biscuiteer:

library(biscuiteer)
## Loading required package: biscuiteerData
## Loading required package: ExperimentHub
## Loading required package: BiocGenerics
## 
## Attaching package: 'BiocGenerics'
## The following objects are masked from 'package:stats':
## 
##     IQR, mad, sd, var, xtabs
## The following objects are masked from 'package:base':
## 
##     anyDuplicated, aperm, append, as.data.frame, basename, cbind,
##     colnames, dirname, do.call, duplicated, eval, evalq, Filter, Find,
##     get, grep, grepl, intersect, is.unsorted, lapply, Map, mapply,
##     match, mget, order, paste, pmax, pmax.int, pmin, pmin.int,
##     Position, rank, rbind, Reduce, rownames, sapply, setdiff, table,
##     tapply, union, unique, unsplit, which.max, which.min
## Loading required package: AnnotationHub
## Loading required package: BiocFileCache
## Loading required package: dbplyr
## Loading biscuiteerData.
## Loading required package: bsseq
## Loading required package: GenomicRanges
## Loading required package: stats4
## Loading required package: S4Vectors
## 
## Attaching package: 'S4Vectors'
## The following object is masked from 'package:utils':
## 
##     findMatches
## The following objects are masked from 'package:base':
## 
##     expand.grid, I, unname
## Loading required package: IRanges
## Loading required package: GenomeInfoDb
## Loading required package: SummarizedExperiment
## Loading required package: MatrixGenerics
## Loading required package: matrixStats
## 
## Attaching package: 'MatrixGenerics'
## The following objects are masked from 'package:matrixStats':
## 
##     colAlls, colAnyNAs, colAnys, colAvgsPerRowSet, colCollapse,
##     colCounts, colCummaxs, colCummins, colCumprods, colCumsums,
##     colDiffs, colIQRDiffs, colIQRs, colLogSumExps, colMadDiffs,
##     colMads, colMaxs, colMeans2, colMedians, colMins, colOrderStats,
##     colProds, colQuantiles, colRanges, colRanks, colSdDiffs, colSds,
##     colSums2, colTabulates, colVarDiffs, colVars, colWeightedMads,
##     colWeightedMeans, colWeightedMedians, colWeightedSds,
##     colWeightedVars, rowAlls, rowAnyNAs, rowAnys, rowAvgsPerColSet,
##     rowCollapse, rowCounts, rowCummaxs, rowCummins, rowCumprods,
##     rowCumsums, rowDiffs, rowIQRDiffs, rowIQRs, rowLogSumExps,
##     rowMadDiffs, rowMads, rowMaxs, rowMeans2, rowMedians, rowMins,
##     rowOrderStats, rowProds, rowQuantiles, rowRanges, rowRanks,
##     rowSdDiffs, rowSds, rowSums2, rowTabulates, rowVarDiffs, rowVars,
##     rowWeightedMads, rowWeightedMeans, rowWeightedMedians,
##     rowWeightedSds, rowWeightedVars
## Loading required package: Biobase
## Welcome to Bioconductor
## 
##     Vignettes contain introductory material; view with
##     'browseVignettes()'. To cite Bioconductor, see
##     'citation("Biobase")', and for packages 'citation("pkgname")'.
## 
## Attaching package: 'Biobase'
## The following object is masked from 'package:MatrixGenerics':
## 
##     rowMedians
## The following objects are masked from 'package:matrixStats':
## 
##     anyMissing, rowMedians
## The following object is masked from 'package:ExperimentHub':
## 
##     cache
## The following object is masked from 'package:AnnotationHub':
## 
##     cache
## 
## 
## 
## Warning: replacing previous import 'BiocParallel::bpstart' by
## 'QDNAseq::bpstart' when loading 'biscuiteer'
orig_bed <- system.file("extdata", "MCF7_Cunha_chr11p15.bed.gz",
                        package="biscuiteer")
orig_vcf <- system.file("extdata", "MCF7_Cunha_header_only.vcf.gz",
                        package="biscuiteer")
bisc <- readBiscuit(BEDfile = orig_bed, VCFfile = orig_vcf,
                    merged = FALSE)
## Checking /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15.bed.gz for import...
## Extracting sample names from /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_header_only.vcf.gz...
## /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15.bed.gz does not have a header. Using VCF file header information to help set column names.
## Assuming unmerged data. Checking now... ...The file might be alright. Double check if you're worried.
## /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15.bed.gz has 254147 indexed loci.
## /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15.bed.gz looks valid for import.
## Reading unmerged input from /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15.bed.gz...
## Excluding CpG sites with uniformly zero coverage...
## Loaded /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15.bed.gz. Creating bsseq object......Done!

Metadata from the biscuit output can be viewed via:

biscuitMetadata(bisc)
## CharacterList of length 3
## [["Reference genome"]] hg19.fa
## [["Biscuit version"]] 0.1.3.20160324
## [["Invocation"]] biscuit pileup -r /primary/vari/genomicdata/genomes/hg19/hg1...

2.3 Combining Methylation Results

In the instance where you have two separate BED files that you would like to analyze in a single bsseq object, you can combine the files using unionize, which is a wrapper around the BiocGenerics function, combine.

shuf_bed <- system.file("extdata", "MCF7_Cunha_chr11p15_shuffled.bed.gz",
                        package="biscuiteer")
shuf_vcf <- system.file("extdata",
                        "MCF7_Cunha_shuffled_header_only.vcf.gz",
                        package="biscuiteer")
bisc2 <- readBiscuit(BEDfile = shuf_bed, VCFfile = shuf_vcf,
                     merged = FALSE)
## Checking /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15_shuffled.bed.gz for import...
## Extracting sample names from /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_shuffled_header_only.vcf.gz...
## /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15_shuffled.bed.gz does not have a header. Using VCF file header information to help set column names.
## Assuming unmerged data. Checking now... ...The file might be alright. Double check if you're worried.
## /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15_shuffled.bed.gz has 254147 indexed loci.
## /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15_shuffled.bed.gz looks valid for import.
## Reading unmerged input from /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15_shuffled.bed.gz...
## Excluding CpG sites with uniformly zero coverage...
## Loaded /tmp/RtmpVb2kuq/Rinst25d1d42665b7fe/biscuiteer/extdata/MCF7_Cunha_chr11p15_shuffled.bed.gz. Creating bsseq object......Done!
comb <- unionize(bisc, bisc2)

2.4 Loading epiBED files

The epiBED file format provides an easy way to analyze read- or fragment-level methylation and genetic information at the same time. readEpibed provides functionality for parsing the RLE strings found in the epiBED file into a GRanges object for analysis in R.

NOTE: The input file must be bgzip’ed and tabix’ed.

epibed.nome <- system.file("extdata", "hct116.nome.epibed.gz", package="biscuiteer")
epibed.bsseq <- system.file("extdata", "hct116.bsseq.epibed.gz", package="biscuiteer")
epibed.nome.gr <- readEpibed(epibed = epibed.nome, genome = "hg19", chr = "chr1")
## Decoding RLE and converting to GRanges
## Collapsing to fragment level
## This will take some time if a large region is being analyzed
epibed.bsseq.gr <- readEpibed(epibed = epibed.bsseq, genome = "hg19", chr = "chr1")
## Decoding RLE and converting to GRanges
## Collapsing to fragment level
## This will take some time if a large region is being analyzed

3 Analysis Functionality

A handful of analysis paths are available in biscuiteer, including A/B compartment inference, age estimation from WGBS data, hypermethylation of Polycomb Repressor Complex (PRC) binding sites, and hypomethylation of CpG-poor “partially methylated domains” (PMDs).

3.1 Inputs for A/B Compartment Inference

When performing A/B compartment inference, the goal is to have something that has roughly gaussian error. getLogitFracMeth uses Dirichlet smoothing to turn raw measurements into lightly moderated, logit-transformed methylated-fraction estimates, which can the be used as inputs to compartmap

reg <- GRanges(seqnames = rep("chr11",5),
               strand = rep("*",5),
               ranges = IRanges(start = c(0,2.8e6,1.17e7,1.38e7,1.69e7),
                                end= c(2.8e6,1.17e7,1.38e7,1.69e7,2.2e7))
              )

frac <- getLogitFracMeth(bisc, minSamp = 1, r = reg)
frac
## GRanges object with 5 ranges and 1 metadata column:
##       seqnames            ranges strand | MCF7_Cunha
##          <Rle>         <IRanges>  <Rle> |  <numeric>
##   [1]    chr11         0-2800000      * |   1.340682
##   [2]    chr11  2800000-11700000      * |   0.575875
##   [3]    chr11 11700000-13800000      * |   1.162989
##   [4]    chr11 13800000-16900000      * |   0.581874
##   [5]    chr11 16900000-22000000      * |   0.442985
##   -------
##   seqinfo: 1 sequence from an unspecified genome; no seqlengths

3.2 Age Estimation

biscuiteer has the functionality to guess the age of the sample(s) provided using the Horvath-style “clock” models (see Horvath, 2013 for more information).

NOTE: The prediction accuracy of this function is entirely dependent on the parameters set by the user. As such, the defaults (as shown in the example below) should only be used as a starting point for exploration by the user.

NOTE: Please cite the appropriate papers for the epigenetic “clock” chosen:

  • For horvath or horvathshrunk
    • Horvath, Genome Biology, 2013
  • For hannum
    • Hannum et al., Molecular Cell, 2013
  • For skinandblood
    • Horvath et al., Aging, 2018
ages <- WGBSage(comb, "horvath")
## Assessing coverage across age-associated regions...
## You have NAs. Change `padding` (15), `minCovg` (5), `useHMMI`, and/or `useENSR`. You have 702 positions in coverage matrix (regions x samples) with less than 5 minCovg. This represents 99.43 % missing data
ages
## $call
## WGBSage(comb, "horvath")
## 
## $droppedSamples
## NULL
## 
## $droppedRegions
## NULL
## 
## $intercept
## [1] 0.6955073
## 
## $methcoefs
## GRanges object with 353 ranges and 3 metadata columns:
##                           seqnames            ranges strand | MCF7_Cunha
##                              <Rle>         <IRanges>  <Rle> |  <numeric>
##      chr1:1168022-1168051     chr1   1168022-1168051      * |         NA
##    chr1:19746550-19746579     chr1 19746550-19746579      * |         NA
##    chr1:23858021-23858050     chr1 23858021-23858050      * |         NA
##    chr1:32084950-32084979     chr1 32084950-32084979      * |         NA
##    chr1:32687553-32687582     chr1 32687553-32687582      * |         NA
##                       ...      ...               ...    ... .        ...
##   chr22:42322132-42322161    chr22 42322132-42322161      * |         NA
##   chr22:43506007-43506036    chr22 43506007-43506036      * |         NA
##   chr22:46449447-46449476    chr22 46449447-46449476      * |         NA
##   chr22:46450093-46450122    chr22 46450093-46450122      * |         NA
##   chr22:50968329-50968358    chr22 50968329-50968358      * |         NA
##                           MCF7_Cunha_shuffled      coefs
##                                     <numeric>  <numeric>
##      chr1:1168022-1168051                  NA  0.6285003
##    chr1:19746550-19746579                  NA  0.0138482
##    chr1:23858021-23858050                  NA -0.1663978
##    chr1:32084950-32084979                  NA  0.0989124
##    chr1:32687553-32687582                  NA  0.0358242
##                       ...                 ...        ...
##   chr22:42322132-42322161                  NA  0.7000011
##   chr22:43506007-43506036                  NA  0.1270524
##   chr22:46449447-46449476                  NA -0.1662689
##   chr22:46450093-46450122                  NA -0.0912389
##   chr22:50968329-50968358                  NA  0.1373155
##   -------
##   seqinfo: 22 sequences from hg19 genome
## 
## $age
##          MCF7_Cunha MCF7_Cunha_shuffled 
##            33.18896            34.88742