B.1 – Introduction to Bioconductor

Martin Morgan

Contents

1 About

1.1 Bioconductor: Analysis and comprehension of high-throughput

genomic data

• Statistical analysis: large data, technological artifacts, designed experiments; rigorous
• Comprehension: biological context, visualization, reproducibility
• High-throughput
  • Sequencing: RNASeq, ChIPSeq, variants, copy number, …
  • Microarrays: expression, SNP, …
  • Flow cytometry, proteomics, images, …

1.2 Packages, vignettes, work flows

• 1296 software packages; also…

  • ‘Annotation’ packages – static data bases of identifier maps, gene models, pathways, etc; e.g., TxDb.Hsapiens.UCSC.hg19.knownGene
  • ’Experiment packages – data sets used to illustrate software functionality, e.g., airway
• Discover and navigate via biocViews

• Package ‘landing page’

  • Title, author / maintainer, short description, citation, installation instructions, …, download statistics
• All user-visible functions have help pages, most with runnable examples
• ‘Vignettes’ an important feature in Bioconductor – narrative documents illustrating how to use the package, with integrated code
• ‘Release’ (every six months) and ‘devel’ branches

• Support site; videos, recent courses

1.3 Package installation and use

• A package needs to be installed once, using the instructions on the package landing page (e.g., DESeq2).

  source("https://bioconductor.org/biocLite.R")
  biocLite(c("DESeq2", "org.Hs.eg.db"))

• biocLite() installs Bioconductor, CRAN, and github packages.

• Once installed, the package can be loaded into an R session

  library(GenomicRanges)

  and the help system queried interactively, as outlined above:

  help(package="GenomicRanges")
  vignette(package="GenomicRanges")
  vignette(package="GenomicRanges", "GenomicRangesHOWTOs")
  ?GRanges

2 Key concepts

2.1 Goals

• Reproducibility
• Interoperability
• Use

2.2 What a few lines of R has to say

x <- rnorm(1000)
y <- x + rnorm(1000)
df <- data.frame(X=x, Y=y)
plot(Y ~ X, df)
fit <- lm(Y ~ X, df)
anova(fit)

## Analysis of Variance Table
## 
## Response: Y
##            Df  Sum Sq Mean Sq F value    Pr(>F)    
## X           1  999.71  999.71  929.56 < 2.2e-16 ***
## Residuals 998 1073.32    1.08                      
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

abline(fit)

2.3 Classes and methods – “S3”

• data.frame()

  • Defines class to coordinate data
  • Creates an instance or object

• plot(), lm(), anova(), abline(): methods defined on generics to transform instances

• Discovery and help

  class(fit)
  methods(class=class(fit))
  methods(plot)
  ?"plot"
  ?"plot.formula"

• tab completion!

2.4 Bioconductor classes and methods – “S4”

• Example: working with DNA sequences

  library(Biostrings)
  dna <- DNAStringSet(c("AACAT", "GGCGCCT"))
  reverseComplement(dna)

  ##   A DNAStringSet instance of length 2
  ##     width seq
  ## [1]     5 ATGTT
  ## [2]     7 AGGCGCC

  data(phiX174Phage)
  phiX174Phage

  ##   A DNAStringSet instance of length 6
  ##     width seq                                            names               
  ## [1]  5386 GAGTTTTATCGCTTCCATGACG...GATTGGCGTATCCAACCTGCA Genbank
  ## [2]  5386 GAGTTTTATCGCTTCCATGACG...GATTGGCGTATCCAACCTGCA RF70s
  ## [3]  5386 GAGTTTTATCGCTTCCATGACG...GATTGGCGTATCCAACCTGCA SS78
  ## [4]  5386 GAGTTTTATCGCTTCCATGACG...GATTGGCGTATCCAACCTGCA Bull
  ## [5]  5386 GAGTTTTATCGCTTCCATGACG...GATTGGCGTATCCAACCTGCA G97
  ## [6]  5386 GAGTTTTATCGCTTCCATGACG...GATTGGCGTATCCAACCTGCA NEB03

  letterFrequency(phiX174Phage, "GC", as.prob=TRUE)

  ##            G|C
  ## [1,] 0.4476420
  ## [2,] 0.4472707
  ## [3,] 0.4472707
  ## [4,] 0.4470850
  ## [5,] 0.4472707
  ## [6,] 0.4470850

• Discovery and help

  class(dna)
  ?"DNAStringSet-class"
  ?"reverseComplement,DNAStringSet-method"

3 High-throughput sequence analysis work flows

• Step 1. Experimental design

  • Simple, replication, sufficient power, covariates and batch effects, …

• Step 2. Wet-lab sequence preparation

  • Figure from http://rnaseq.uoregon.edu/

• Step 3. (Illumina) Sequencing

  • Bentley et al., 2008, doi:10.1038/nature07517
  • Primary output: FASTQ files of short reads and their quality scores.

• Step 4. Alignment

  • Choose to match task, e.g., Rsubread, Bowtie2 good for ChIPseq, some forms of RNAseq; BWA, GMAP better for variant calling
  • Primary output: BAM files of aligned reads
  • More recently: kallisto and similar programs that produce tables of reads aligned to transcripts

• Step 5. Reduction

  • e.g., RNASeq ‘count table’ (simple spreadsheets), DNASeq called variants (VCF files), ChIPSeq peaks (BED, WIG files)

• Step 6. Analysis

  • Differential expression, peak identification, differential binding, …

• Step 7. Comprehension

  • Biological context; annotation, gene set analysis, …

4 Bioconductor sequencing ecosystem

Alt Sequencing Ecosystem