[BioC] 'SRAdb useful for visualizing BAM files'

[Sean Davis]

On Sun, Apr 11, 2010 at 12:41 PM, Aaron Mackey <ajmackey at gmail.com> wrote:
> Right, I understand that IGV is capable of this -- what I'm not clear about
> is whether the NCBI's SRA is capable of "serving" the BAM files in this
> fashion, and if the R SRAdb package is capable of managing these forms of
> connections, or must I still download all the BAMs (via SRAdb or otherwise).
>

Hi, Aaron.  NCBI SRA does not serve BAM files at all, only fastq
files.  The functionality of interfacing with IGV is not directly
related to the SRA querying facilities, really.  SRAdb IGV interfacing
is through filenames or URLs; no connections are really managed in R
to make this happen.  Only filenames or URLs are passed via an http
protocol from R to IGV.  The rest is done by IGV.  Since IGV is quite
capable of reading from a URL in a random-access fashion, BAM files on
an appropriate server can be read without downloading.

Hope that helps,
Sean

> On Sat, Apr 10, 2010 at 10:42 AM, Jack Zhu <zhujack at mail.nih.gov> wrote:
>>
>> > ---------- Forwarded message ----------
>> > From: Aaron Mackey <ajmackey at gmail.com>
>> > Date: Sat, Apr 10, 2010 at 7:49 AM
>> > Subject: Re: [BioC] 'SRAdb useful for visualizing BAM files'
>> > To: Sean Davis <seandavi at gmail.com>
>> > Cc: Paul Shannon <pshannon at systemsbiology.org>, bioc
>> > <bioconductor at stat.math.ethz.ch>
>> >
>> >
>> > Thanks for the info Sean.  Will this use the remote-BAM access
>> > capabilities of IGV, or does it entail downloading the entire BAM from
>> > SRA before viewing?
>>
>> Yes, IGV has capabilities of using http random access, which means
>> data is dynamically loaded into IGV viewing window without loading an
>> entire BAM file in advance.  Please see details here:
>> http://www.broadinstitute.org/igv/ControlIGV .  From our experience,
>> the random data access (Google Map like) in IGV make data loading
>> faster and loading large multiple tracks possible without huge memory
>> problem.
>>
>> Example using an URL directly:
>>
>>
>> http://www.broadinstitute.org/igv/dynsession/igv.jnlp?user=anonymous&sessionURL=http://watson.nci.nih.gov/~zhujack/SRAdb_bam/SRR036007.chr1.hg18.bwa.sorted.bam&locus=chr1:1-1000000
>>
>>   This will launch a new IGV and access the bam from an URL.
>>
>> Example using the SRAdb package:
>>
>>        library(SRAdb)
>>        sra_dbname <- 'SRAmetadb.sqlite'
>>        sra_con <- dbConnect(dbDriver("SQLite"), sra_dbname)
>>
>>        ## If  there is no IGV running in your computer, the following
>> statement will launch a new IGV and load BAM files to it:
>>        files = list.files(system.file('extdata', package = "SRAdb"),
>> pattern='.bam$', full.names=TRUE)
>>        load2newIGV(files, locus='chr1:1-1000000')
>>
>> Please let me know if you have any questions.
>>
>> Jack
>>
>>
>>
>>
>> >
>> > -Aaron
>> >
>> > On Fri, Apr 9, 2010 at 10:21 PM, Sean Davis <seandavi at gmail.com> wrote:
>> >>
>> >> On Fri, Apr 9, 2010 at 9:40 PM, Paul Shannon
>> >> <pshannon at systemsbiology.org> wrote:
>> >> > Martin suggests  ("An Introduction to Rsamtools", April 8th, 2010):
>> >> >
>> >> >   'Users might also find the interface to the integrative genome
>> >> > browser (IGV) in SRAdb useful for visualizing BAM files.'
>> >> >
>> >> > I have tried googling many versions of the relevant words, and cannot
>> >> > find anything about this.  Is it an R interface to IGV?  Any ideas about how
>> >> > I can find out more?
>> >> >
>> >> > Thanks!
>> >>
>> >> Hi, Paul.
>> >>
>> >> SRAdb is an R package for searching the NCBI SRA database in a more
>> >> user-friendly way and for accessing the data in SRA.  One of the
>> >> facilities that is included is an interface to start IGV and load BAM
>> >> files of interest directly from R.  The package has been submitted to
>> >> bioconductor and is in review.  However, if you like, you can
>> >> potentially install it now (assuming that you can deal with source
>> >> package installation) by doing:
>> >>
>> >>
>> >> install.packages('SRAdb',repos='http://watson.nci.nih.gov/R/repos',type='source')
>> >>
>> >> There is a vignette included.
>> >>
>> >> Sean
>> >>
>> >> _______________________________________________
>> >> Bioconductor mailing list
>> >> Bioconductor at stat.math.ethz.ch
>> >> https://stat.ethz.ch/mailman/listinfo/bioconductor
>> >> Search the archives:
>> >> http://news.gmane.org/gmane.science.biology.informatics.conductor
>> >
>
>