[Bioperl-l] Database Retrieval

[Sean Davis]

On 8/7/06 10:41 AM, "Sean Davis" <sdavis2 at mail.nih.gov> wrote:

> 
> 
> 
> On 8/7/06 9:58 AM, "Rick Scavetta" <RScavett at uni-koeln.de> wrote:
> 
>> Hello Programmers,
>> 
>> I have a list of mouse GeneIDs for which I have extracted the RefSeqs for.
>> With these accession numbers I want to know what are the three closest
>> upstream and downstream genes (and orientation, if possible) to my gene of
>> interest. Is these some way of finding this out? Any suggestions?
> 
> I would look at using the UCSC genome browser data.  You can download the
> refGene table, order it by chromosome and start, then do whatever
> manipulation you like.  Your problem can probably be solved with perl, but
> if you have experience using R (the statistics software), using it may be
> more straightforward.  You could also potentially do this using SQL queries
> of the UCSC MySQL database.
> 
> Is there a reason to use "number of genes"?  If you can simply restrict to a
> number of base pairs, the problem can (probably) be solved using only the
> UCSC table browser.
> 
>> Also, I would also like to know something about the expression of a
>> particular gene of interest. e.g. ba querying the Novartis Gene Atlas
>> (http://symatlas.gnf.org/SymAtlas/). Is there a module for handling
>> submissions and retrieving results of this sort?
> 
> No, but, again, UCSC hosts these data and they have nice query tools, a
> publicly-accessible MySQL database containing the information, and
> downloadable tables for the data.

Oh, and I should have mentioned that NCBI GEO hosts these data, which can be
downloaded as a "spreadsheet" from here:

http://www.ncbi.nlm.nih.gov/geo/gds/gds_browse.cgi?gds=596

Sean