[BioC] biomaRt error in getGene

[Georg Otto]

Hi,

I have a problem using biomaRt. I want to retrieve information connected to a probe, and do something like this:

> mart<-useMart("ensembl", dataset="drerio_gene_ensembl")
Checking attributes and filters ... ok
> getAffyArrays(mart)
[1] "affy_zebrafish"
> getGene(id="Dr.1.1.S1_at", array="affy_zebrafish", mart=mart)
Error in getBM(attributes = c(attrib, "hgnc_symbol", "description", chrname,  : 
	attribute: hgnc_symbol not found, please use the function 'listAttributes' to get valid attribute names


> listAttributes(mart=mart)

<snip>
  [8] "adf_swissprot"                                        
  [9] "affy_zebrafish"                                       
 [10] "affy_zebrafish_primary_db"                            
 [11] "agilent_g2518a"                                       
<snip>

So it seems that the attribute "affy_zebrafish" exists.

What is wrong here? I had this problem before, and I got a reply from
Steffen Durinck (see below) that it has to do with an inconistency in
attribute and filter naming, i.e. the attribute for the affyids was
zebrafish_affy and the filter was called affy_zebrafish. I was told
that this will be fixed in the next ensembl release. It seems that the
diffrence between the array and the attribute has been repaired, since
both are now called affy_zebrafish, but the problem still persists.

> sessionInfo()
Version 2.3.1 (2006-06-01) 
powerpc-apple-darwin8.6.0 

attached base packages:
[1] "methods"   "stats"     "graphics"  "grDevices" "utils"     "datasets" 
[7] "base"     

other attached packages:
     DBI  biomaRt    RCurl      XML 
"0.1-10"  "1.6.0"  "0.6-2" "0.99-7"

Cheers,

Georg



> Hi,
>
>> My understanding is that the BioMart folks are making changes to the
>> table names for the BioMart database, and this is happening right now
>> (e.g, right after BioC 1.8 is released). Unfortunately this means that
>> some of the convenience functions like getGene() are being broken.


>
> This is correct however here there is a problem with the zebrafish dataset
> as well.  Part of the error here is produced by an inconistency in
> attribute and filter naming.  The attribute for the affyids is
> zebrafish_affy and the filter is called affy_zebrafish.  The getGene
> function expects these to have the same name and thus generates an error.
>
> To make the BioMart datasets better we can post these inconsistencies to
> the corresponding BioMart mailinglist (in this case the Ensembl helpdesk)
> so they get fixed in the next database release.
>
> Best,
> Steffen
>
>> Hi Georg,
>>
>> Georg Otto wrote:
>>> Hi,
>>>
>>> using biomaRt, I get an error:
>>>
>>>
>>>>mart<-useMart("ensembl", dataset= "drerio_gene_ensembl")
>>>>getAffyArrays(mart)
>>>
>>> [1] "affy_zebrafish"
>>>
>>>>getGene(id=genes.regulated.mas5, array="affy_zebrafish" , mart=mart)
>>>
>>> Error in getBM(attributes = c(attrib, "hgnc_symbol", "description",
>>> chrname,  :
>>>         attribute: affy_zebrafish not found, please use the function
>>> 'listAttributes' to get valid attribute names
>>>
>>>
>>> Then I use listAttributes() as requested:
>>>
>>>
>>>>listAttributes(mart=mart)
>>>
>>>
>>> and get the following output:
>>>
>>> <snip>
>>> [309] "zebrafish_affy"
>>> [310] "zebrafish_affy_primary_db"
>>> <snip>
>>>
>>> using "zebrafish_affy" instead of "affy_zebrafish" does not help,
>>> however:
>>>
>>>
>>>>getGene(id=genes.regulated.mas5, array="zebrafish_affy" , mart=mart)
>>>
>>> Error in getBM(attributes = c(attrib, "hgnc_symbol", "description",
>>> chrname,  :
>>>         attribute: hgnc_symbol not found, please use the function
>>> 'listAttributes' to get valid attribute names
>>>
>>> Any hint will be appreciated.
>>
>> My understanding is that the BioMart folks are making changes to the
>> table names for the BioMart database, and this is happening right now
>> (e.g, right after BioC 1.8 is released). Unfortunately this means that
>> some of the convenience functions like getGene() are being broken.
>>
>> I think your best bet is to use getBM() directly, and query for things
>> that you see when you do listAttributes(mart).
>>
>> HTH,
>>
>> Jim