Why do we use ranges? The terrific Science Web summarizes: Most bioinformaticians to be replaced by BEDTools

While the simple linear sequence representation of genomes begins to break down when we consider structural variants like insertions, deletions, translocations, copy number variants, etc. -- everything is better thought of as a graph. There's exciting new work in representing and working with sequences as graphs, e.g. the preprints Dilthey et al., 2014 and Paten et al., 2014.

As of writing this chapter, parts of the IRanges package are being split off into the new S4Vectors package. I only discuss these lower-level topics briefly, but if you want more information see the IRanges vignette and the S4Vectors page. I don't link these directly in the book because these links and vignette content may change.

• Mus_musculus.GRCm38.75_chr1.gtf.gf are chromosome 1 annotations extracted from Mus_musculus.GRCm38.75.gtf.gz downloaded from Ensembl's FTP (ftp://ftp.ensembl.org/pub/release-75/gtf/mus_musculus) on 2014-08-02.

    gzcat Mus_musculus.GRCm38.75.gtf.gz | egrep "^(1\t|#)" | gzip > Mus_musculus.GRCm38.75_chr1.gtf.gf
  

• mm10_snp137_chr1_trunc.bed.gz is a randomly sampled, shorter version of mm10_snp137_chr1.bed.gz (since Github doesn't play well with large files). mm10_snp137_chr1.bed.gz was downloaded from the UCSC Genome Browser's Table Browser (http://genome.ucsc.edu/cgi-bin/hgTables). This was file was downloaded on 2014-08-01. The truncated file was created with (you need GNU sort for this):

    $ gzcat mm10_snp137_chr1.bed.gz | sort --random-sort | head -n 2700000 | gzip > mm10_snp137_chr1_trunc.bed.gz
  

• Mus_musculus.GRCm38_genome.txt is a tab-delimited file of all chromosome lengths from the mm10/GRCm38 genome version. It was created with:

    curl ftp://ftp.ensembl.org/pub/release-75/fasta/mus_musculus/dna/Mus_musculus.GRCm38.75.dna_rm.toplevel.fa.gz \
    | bioawk -c fastx '{print $name"\t"length($seq)}' > Mus_musculus.GRCm38_genome.txt
  

I use Python and R's indexing to introduce 0- and 1-based range systems. Here's Python's creator Guido van Rossum on why Python uses 0-based indexing.

Why use biocLite() rather than install.packages()? See this description from Bioconductor.