[Bioc-devel] how to get genomic sequences?

[Sean Davis]

Roger Liu wrote:
> Hi,
>
> I have a set of data with chromosome number and coordinates of the sequences
> such as,chr10, start 1000, end 2000.
> I have tried using biomart to retrieve the genomic sequences for my dataset,
> but I didn't get success, I used seqType argument as:
> seqType="genomic", it reported error as"The type of sequence specified with
> seqType is not available. Please select from: cdna, peptide, 3utr, 5utr",
> but I have seen this "genomic" argument for seqType in the help file. So
> what's going on there?
>
> Or anyone can recommend a package which can help me retrieve the genomic
> sequences from hg18 with known chromosome number and sequences
> coordinates(start and end).
>   
URLs like so:

http://genome.ucsc.edu/cgi-bin/das/hg18/dna?segment=chr1:1000,2000;segment=chr2:1000,2000

will get you sequences of interest.  These come as simple well-formed 
XML, so you can use the XML package to parse them pretty easily.  You 
can do one-at-a-time, or string together a bunch.

Sean