Hi Andreas,

The reason you are not getting all of the fields in the BED result is 
that (I think) you are starting with the knownCanonical table, which 
doesn't include the 9 optional fields.  If you start with the knownGene 
table and then filter the results so that only the canonical genes are 
included, you will get all 12 fields in the output.

To filter on the canonical genes, first get a list of the canonical 
identifiers by selecting knownCanonical in the Table Browser, and then 
"selected fields in primary and related tables."  Make sure "region: 
genome" is selected, hit "get output," and on the next page choose the 
"transcript" field.  Now copy that list (there should be 31,227 
knownCanonical identifiers for hg19) and go back to the main Table 
Browser page.

Select the knownGene table this time, and then hit "filter: create". 
Paste the whole list of canonical identifiers into the "name does match" 
box.  (I know it doesn't look like it could handle thirty thousand 
identifiers, but it does!)  Choose BED as the output format and hit "get 
output."  You should get a BED 12 of the canonical UCSC Genes.

When I tried this here, the Table Browser was a bit slow the first time 
and eventually gave an error, but when I hit the back button in my 
browser and tried it again, I got the results successfully.  Let us know 
if the procedure I describe doesn't work.

If you have further questions, please contact us again at 
[email protected].

--
Brooke Rhead
UCSC Genome Bioinformatics Group


On 5/9/12 9:46 AM, andreas schueler wrote:
> Hi,
> I tried to get a table of the canonical UCSC genes from the hg19 assembly
> in BED12 format from the table browser (with
> the blockstarts and blocksizes fields corresponding to exons).
> When I select BED as output format, I get BED3 data - I tried chosing
> "selected fields from primary and related tables"
> as output format and I can find almost all the data for the 9 optional BED
> fields in the hg19.refGene table, but not for the
> blocksizes field (exon end positions are there, but not exon sizes).
> Is there an easy way to get BED12 formatted data from the table browser ?
> cheers,
> Andreas
>
_______________________________________________
Genome maillist  -  [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to