Lecture 2 – Introduction to Bioconductor

Martin Morgan

Contents

1 History

• Started 2002 as a platform for understanding analysis of microarray data
• More than 1,300 packages. Domains of expertise:
  • Sequencing (RNASeq, ChIPSeq, single-cell, called variants, …)
  • Microarrays (methylation, expression, copy number, …)
  • flow cytometry
  • proteomics
  • …
• Important themes
  • Reproducible research
  • Interoperability between packages & work kflows
  • Usability

2 High-throughput sequence work flow

3 DNA Sequences

library(Biostrings)

dna <- DNAStringSet(c("AAACTG", "CCCTTCAAC", "TACGAA"))
dna

##   A DNAStringSet instance of length 3
##     width seq
## [1]     6 AAACTG
## [2]     9 CCCTTCAAC
## [3]     6 TACGAA

length(dna)

## [1] 3

dna[c(1, 3, 1)]

##   A DNAStringSet instance of length 3
##     width seq
## [1]     6 AAACTG
## [2]     6 TACGAA
## [3]     6 AAACTG

width(dna)

## [1] 6 9 6

reverseComplement(dna)

##   A DNAStringSet instance of length 3
##     width seq
## [1]     6 CAGTTT
## [2]     9 GTTGAAGGG
## [3]     6 TTCGTA

Biostrings themes

• Valid data, e.g., alphabet
• ‘Vector’ interface: length(), [, …
• Specialized operations, e.g,. reverseComplement()

Help!

methods(class="DNAStringSet")
?"DNAStringSet"
browseVignettes(package="Biostrings")

4 Genomic Ranges

library(GenomicRanges)

gr <- GRanges(c("chr1:100-120", "chr1:115-130", "chr2:200-220"))
gr

## GRanges object with 3 ranges and 0 metadata columns:
##       seqnames     ranges strand
##          <Rle>  <IRanges>  <Rle>
##   [1]     chr1 [100, 120]      *
##   [2]     chr1 [115, 130]      *
##   [3]     chr2 [200, 220]      *
##   -------
##   seqinfo: 2 sequences from an unspecified genome; no seqlengths

shift(gr, 1)

## GRanges object with 3 ranges and 0 metadata columns:
##       seqnames     ranges strand
##          <Rle>  <IRanges>  <Rle>
##   [1]     chr1 [101, 121]      *
##   [2]     chr1 [116, 131]      *
##   [3]     chr2 [201, 221]      *
##   -------
##   seqinfo: 2 sequences from an unspecified genome; no seqlengths

reduce(gr)

## GRanges object with 2 ranges and 0 metadata columns:
##       seqnames     ranges strand
##          <Rle>  <IRanges>  <Rle>
##   [1]     chr1 [100, 130]      *
##   [2]     chr2 [200, 220]      *
##   -------
##   seqinfo: 2 sequences from an unspecified genome; no seqlengths

anno <- GRanges(c("chr1:110-150", "chr2:150-210"))
countOverlaps(anno, gr)

## [1] 2 1

GenomicRanges

• Data (e.g., aligned reads, called peaks, copy number)
• Annotations (e.g., genes, exons, transcripts)
• Close relation to BED files (see rtracklayer::import.bed() and HelloRanges)
• Also vector interface – length(), [, etc.

Intra-range operations

• e.g., shift(), narrow(), flank()

Inter-range operations

• e.g., reduce(), coverage(), gaps(), disjoin()

Between-range operations

• e.g., countOverlaps(), findOverlaps(), summarizeOverlaps()

Help!

methods(class="GRanges")
methods(class="GRangesList")
?"GRanges"
?"GRangesList"
browseVignettes(package="GenomicRanges")

4.1 Lists of Genomic Ranges

• e.g., exons-within-transcripts, alignments-within-reads

5 Summarized Experiments