Bioconductor can import diverse sequence-related file types, including fasta, fastq, BAM, VCF, gff, bed, and wig files, among others. Packages support common and advanced sequence manipulation operations such as trimming, transformation, and alignment. Domain-specific analyses include quality assessment, ChIP-seq, differential expression, RNA-seq, and other approaches. Bioconductor includes an interface to the Sequence Read Archive (via the SRAdb package).
This workflow walks through the annotation of a generic set of ranges with Bioconductor packages. The ranges can be any user-defined region of interest or can be from a public file.
As a first step, data are put into a GRanges object so we can take advantage of overlap operations and store identifiers as metadata columns.
The first set of ranges are variants from a dbSNP Variant Call Format (VCF) file. This file can be downloaded from the ftp site at NCBI ftp://ftp.ncbi.nlm.nih.gov/snp/ and imported with readVcf() from the VariantAnnotation package. Alternatively, the file is available as a pre-parsed VCF object in the AnnotationHub.
library(VariantAnnotation)
library(AnnotationHub)
The Hub returns a VcfFile object with a reference to the file on disk.
hub <- AnnotationHub()
fl <- hub[['AH47004']]
fl
## class: VcfFile
## path: /var/lib/jenkins/.AnnotationHub/52446
## index: /var/lib/jenkins/.AnnotationHub/52447
## isOpen: FALSE
## yieldSize: NA
Read the data into a VCF object:
vcf <- readVcf(fl, "hg19")
dim(vcf)
## [1] 114699 0
Overlap operations require that seqlevels and the genome of the objects match. Here we modify the VCF seqlevels to match the TxDb.
library(TxDb.Hsapiens.UCSC.hg19.knownGene)
txdb_hg19 <- TxDb.Hsapiens.UCSC.hg19.knownGene
head(seqlevels(txdb_hg19))
## [1] "chr1" "chr2" "chr3" "chr4" "chr5" "chr6"
seqlevels(vcf)
## [1] "1" "2" "3" "4" "5" "6" "7" "8" "9" "10" "11" "12" "13" "14"
## [15] "15" "16" "17" "18" "19" "20" "21" "22" "X" "Y" "MT"
seqlevels(vcf) <- paste0("chr", seqlevels(vcf))
Sanity check to confirm we have matching seqlevels.
intersect(seqlevels(txdb_hg19), seqlevels(vcf))
## [1] "chr1" "chr2" "chr3" "chr4" "chr5" "chr6" "chr7" "chr8"
## [9] "chr9" "chr10" "chr11" "chr12" "chr13" "chr14" "chr15" "chr16"
## [17] "chr17" "chr18" "chr19" "chr20" "chr21" "chr22" "chrX" "chrY"
The genomes already match so no change is needed.
unique(genome(txdb_hg19))
## [1] "hg19"
unique(genome(vcf))
## [1] "hg19"
We are only interested in the standard chromosomes so we drop the rest.
txdb_hg19 <- keepStandardChromosomes(txdb_hg19)
vcf <- keepStandardChromosomes(vcf)
The GRanges in a VCF object is extracted with 'rowRanges()'.
gr_hg19 <- rowRanges(vcf)
The second set of ranges is a user-defined region of chromosome 4 in mouse. The idea here is that any region, known or unknown, can be annotated with the following steps.
Load the TxDb package and keep only the standard chromosomes.
library(TxDb.Mmusculus.UCSC.mm10.ensGene)
txdb_mm10 <- keepStandardChromosomes(TxDb.Mmusculus.UCSC.mm10.ensGene)
We are creating the GRanges from scratch and can specify the seqlevels (chromosome names) to match the TxDb.
head(seqlevels(txdb_mm10))
## [1] "chr1" "chr2" "chr3" "chr4" "chr5" "chr6"
gr_mm10 <- GRanges("chr4", IRanges(c(4000000, 107889000), width=1000))
Now assign the genome.
unique(genome(txdb_mm10))
## [1] "mm10"
genome(gr_mm10) <- "mm10"
locateVariants() in the VariantAnnotation package annotates ranges with transcript, exon, cds and gene ID's from a TxDb. Various extractions are performed on the TxDb (exonsBy(), transcripts(), cdsBy(), etc.) and the result is overlapped with the ranges. An appropriate GRangesList can also be supplied as the annotation. Different variants such as 'coding', 'fiveUTR', 'threeUTR', 'spliceSite', 'intron', 'promoter', and 'intergenic' can be searched for by passing the appropriate constructor as the 'region' argument. See ?locateVariants for details.
loc_hg19 <- locateVariants(gr_hg19, txdb_hg19, AllVariants())
table(loc_hg19$LOCATION)
##
## spliceSite intron fiveUTR threeUTR coding intergenic
## 311 209819 1276 4933 11383 48467
## promoter
## 12440
loc_mm10 <- locateVariants(gr_mm10, txdb_mm10, AllVariants())
table(loc_mm10$LOCATION)
##
## spliceSite intron fiveUTR threeUTR coding intergenic
## 6 1 0 0 0 0
## promoter
## 12
The ID's returned from locateVariants() can be used in select() to map to ID's in other annotation packages.
library(org.Hs.eg.db)
cols <- c("UNIPROT", "PFAM")
keys <- na.omit(unique(loc_hg19$GENEID))
head(select(org.Hs.eg.db, keys, cols, keytype="ENTREZID"))
## ENTREZID UNIPROT PFAM
## 1 9636 P05161 PF00240
## 2 375790 O00468 PF00008
## 3 375790 O00468 PF00050
## 4 375790 O00468 PF00053
## 5 375790 O00468 PF00054
## 6 375790 O00468 PF01390
The 'keytype' argument specifies that the mouse TxDb contains Ensembl instead of Entrez gene id's.
library(org.Mm.eg.db)
keys <- unique(loc_mm10$GENEID)
head(select(org.Mm.eg.db, keys, cols, keytype="ENSEMBL"))
## ENSEMBL UNIPROT PFAM
## 1 ENSMUSG00000028236 Q7TQA3 PF00106
## 2 ENSMUSG00000028608 Q8BHG2 PF05907
Files stored in the AnnotationHub have been pre-processed into ranged-based R objects such as a GRanges, GAlignments and VCF. The positions in our GRanges can be overlapped with the ranges in the AnnotationHub files. This allows for easy subsetting of multiple files, resulting in only the ranges of interest.
Create a 'hub' from AnnotationHub and filter the files based on organism and genome build.
hub <- AnnotationHub()
hub_hg19 <- subset(hub,
(hub$species == "Homo sapiens") & (hub$genome == "hg19"))
length(hub_hg19)
## [1] 5539
Iterate over the first 3 files and extract ranges that overlap 'gr_hg19'.
ov_hg19 <- lapply(1:3, function(i) subsetByOverlaps(hub_hg19[[i]], gr_hg19))
Inspect the results.
names(ov_hg19) <- names(hub_hg19)[1:3]
lapply(ov_hg19, head, n=3)
## $AH3166
## GRanges object with 3 ranges and 5 metadata columns:
## seqnames ranges strand |
## <Rle> <IRanges> <Rle> |
## [1] chr14 [ 23388231, 23388425] - |
## [2] chr11 [118436595, 118436793] - |
## [3] chr7 [141438132, 141438281] + |
## name score level
## <character> <integer> <numeric>
## [1] chr14:23388231:23388425:-:1.08:0.993703 0 370.1525
## [2] chr11:118436595:118436793:-:2.21:0.999420 0 255.1222
## [3] chr7:141438132:141438281:+:0.37:0.999969 0 197.9053
## signif score2
## <numeric> <integer>
## [1] 5.15e-09 0
## [2] 8.41e-09 0
## [3] 9.31e-09 0
## -------
## seqinfo: 24 sequences from hg19 genome
##
## $AH3912
## GRanges object with 3 ranges and 5 metadata columns:
## seqnames ranges strand | name score signalValue
## <Rle> <IRanges> <Rle> | <character> <integer> <numeric>
## [1] chr1 [948050, 950479] * | . 0 425.73300
## [2] chr1 [954433, 956440] * | . 0 107.74700
## [3] chr1 [970635, 971318] * | . 0 2.16979
## pValue qValue
## <numeric> <numeric>
## [1] 324.00000 -1
## [2] 324.00000 -1
## [3] 1.63691 -1
## -------
## seqinfo: 23 sequences from hg19 genome
##
## $AH3913
## GRanges object with 3 ranges and 6 metadata columns:
## seqnames ranges strand | name score
## <Rle> <IRanges> <Rle> | <character> <integer>
## [1] chr1 [1048860, 1049010] * | . 0
## [2] chr1 [3838920, 3839070] * | . 0
## [3] chr1 [6051680, 6051830] * | . 0
## signalValue pValue qValue peak
## <numeric> <numeric> <numeric> <integer>
## [1] 60 21.9043 -1 -1
## [2] 451 324.0000 -1 -1
## [3] 107 321.6620 -1 -1
## -------
## seqinfo: 23 sequences from hg19 genome
Annotating the mouse ranges in the same fashion is left as an exercise.
For the set of dbSNP variants that fall in coding regions, amino acid changes can be computed. The output contains one line for each variant-transcript match which can result in multiple lines for each variant.
library(BSgenome.Hsapiens.UCSC.hg19)
head(predictCoding(vcf, txdb_hg19, Hsapiens), 3)
## GRanges object with 3 ranges and 17 metadata columns:
## seqnames ranges strand | paramRangeID
## <Rle> <IRanges> <Rle> | <factor>
## rs397514721 chr1 [1233203, 1233203] - | <NA>
## rs397514721 chr1 [1233203, 1233203] - | <NA>
## rs397514721 chr1 [1233203, 1233203] - | <NA>
## REF ALT QUAL FILTER
## <DNAStringSet> <DNAStringSetList> <numeric> <character>
## rs397514721 T A <NA> .
## rs397514721 T A <NA> .
## rs397514721 T A <NA> .
## varAllele CDSLOC PROTEINLOC QUERYID
## <DNAStringSet> <IRanges> <IntegerList> <integer>
## rs397514721 T [ 317, 317] 106 49
## rs397514721 T [1001, 1001] 334 49
## rs397514721 T [1127, 1127] 376 49
## TXID CDSID GENEID CONSEQUENCE
## <character> <IntegerList> <character> <factor>
## rs397514721 4150 12185 116983 nonsynonymous
## rs397514721 4151 12185 116983 nonsynonymous
## rs397514721 4152 12185 116983 nonsynonymous
## REFCODON VARCODON REFAA VARAA
## <DNAStringSet> <DNAStringSet> <AAStringSet> <AAStringSet>
## rs397514721 GAG GTG E V
## rs397514721 GAG GTG E V
## rs397514721 GAG GTG E V
## -------
## seqinfo: 24 sequences from hg19 genome; no seqlengths
sessionInfo()
## R version 3.2.0 (2015-04-16)
## Platform: x86_64-unknown-linux-gnu (64-bit)
## Running under: Ubuntu precise (12.04.4 LTS)
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=C
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] stats4 parallel stats graphics grDevices utils datasets
## [8] methods base
##
## other attached packages:
## [1] BSgenome.Hsapiens.UCSC.hg19_1.4.0
## [2] BSgenome_1.36.1
## [3] rtracklayer_1.28.5
## [4] org.Mm.eg.db_3.1.2
## [5] org.Hs.eg.db_3.1.2
## [6] RSQLite_1.0.0
## [7] DBI_0.3.1
## [8] TxDb.Mmusculus.UCSC.mm10.ensGene_3.1.2
## [9] TxDb.Hsapiens.UCSC.hg19.knownGene_3.1.2
## [10] GenomicFeatures_1.20.1
## [11] AnnotationDbi_1.30.1
## [12] Biobase_2.28.0
## [13] AnnotationHub_2.0.2
## [14] VariantAnnotation_1.14.3
## [15] Rsamtools_1.20.4
## [16] Biostrings_2.36.1
## [17] XVector_0.8.0
## [18] GenomicRanges_1.20.5
## [19] GenomeInfoDb_1.4.1
## [20] IRanges_2.2.4
## [21] S4Vectors_0.6.0
## [22] BiocGenerics_0.14.0
##
## loaded via a namespace (and not attached):
## [1] Rcpp_0.11.6 BiocInstaller_1.18.3
## [3] formatR_1.2 futile.logger_1.4.1
## [5] bitops_1.0-6 futile.options_1.0.0
## [7] tools_3.2.0 zlibbioc_1.14.0
## [9] biomaRt_2.24.0 digest_0.6.8
## [11] evaluate_0.7 shiny_0.12.1
## [13] stringr_1.0.0 httr_0.6.1
## [15] knitr_1.10.5 R6_2.0.1
## [17] XML_3.98-1.2 BiocParallel_1.2.5
## [19] lambda.r_1.1.7 magrittr_1.5
## [21] htmltools_0.2.6 GenomicAlignments_1.4.1
## [23] mime_0.3 interactiveDisplayBase_1.6.0
## [25] xtable_1.7-4 httpuv_1.3.2
## [27] stringi_0.4-1 RCurl_1.95-4.6
Exercise 1: VCF header and reading data subsets.
VCF files can be large and it's often the case that only a subset of variables or genomic positions are of interest. The scanVcfHeader() function in the VariantAnnotation package retrieves header information from a VCF file. Based on the information returned from scanVcfHeader() a ScanVcfParam() object can be created to read in a subset of data from a VCF file.
Exercise 2: Annotate the mouse ranges in 'gr_mm10' with AnnotationHub files.
Exercise 3: Annotate a gene range from Saccharomyces Scerevisiae.
[ Back to top ]