[Bioperl-l] locuslink - gene positions?

[Stefan Kirov]

Hi Sean,
Thanks! Well, honestly speaking I think the problem is NCBI. They change 
formats, file locations frequently enough to keep us busy. Now they move 
from LocusLink to Entrez Gene, so some stuff is vanishing from LL_tmpl 
and moving to the new ftp dir in a separate file. So we have to write 
new parser, etc. Homologene also is changing the format and this will 
lead to a schema change. MGI map is also not in LL_tmpl so we have to 
get this from  jax. And so forth...
I think H_inv is a nice idea. Another idea we are exploring is Jaspar...
Stefan

Sean Davis wrote:

> Stefan,
>
> Your database design is quite nice!  I'm anxious to try it out.  What 
> are the "update issues"?
>
> We have microarray spot annotation that we have built for custom 
> arrays.  Currently, I link to home-brew tables with annotation 
> information of interest to our users, but it would certainly be nice 
> to have a normalized version of the information.  Also, I would be 
> interested in putting the H-inv data into the schema at some point, as 
> we mainly work in human.
>
> Sean
>
> On Aug 25, 2004, at 5:01 PM, Stefan Kirov wrote:
>
>> I think you are right- there are no coordinates and there should be 
>> (contig coordinates of course). Actually the contig accession number 
>> is in the object as a DBLink object. I don't think DBLink is a 
>> suitable place for the contig coordinates. Another problem with LLink 
>> module is that it permits null values for the primary ID for sime 
>> DBLink objects (LLID 218 for examples). This is definitely a bug.
>> I guess you can request this to be added to the module- file a bug 
>> report. I can probably take a look at that but not today, sorry.
>> You can also check Ensembl API- it may be more straightforward for 
>> your needs. Something that I am doing (this is not an advertisement) 
>> might be also helpful: genereg.ornl.gov/gkdb, but we are struggling 
>> with some update issues, so I wouldn't use it right now.
>> Have fun,
>> Stefan
>>
>> Fontaine, Burr R wrote:
>>
>>> I am trying to mirror portions of Locuslink and have figured out how 
>>> to get all of the data elements I need from the locuslink module 
>>> except for the gene starting postion (in bp coordinates). Chromosome 
>>> coordinates would be best but contig coordinates are OK if there is 
>>> a way to convert them. I've looked at the location modules but don't 
>>> see any obvious answers there. Can someone help me figure out the 
>>> best way to do this?
>>>
>>> Thanks in advance for your help.
>>>
>>> Burr Fontaine
>>>
>>> _______________________________________________
>>> Bioperl-l mailing list
>>> Bioperl-l at portal.open-bio.org
>>> http://portal.open-bio.org/mailman/listinfo/bioperl-l
>>>
>>
>> _______________________________________________
>> Bioperl-l mailing list
>> Bioperl-l at portal.open-bio.org
>> http://portal.open-bio.org/mailman/listinfo/bioperl-l
>
>