[ensembl-dev] $protein_tree_adaptor->fetch_all_roots

[Matthieu Muffato]

We have indeed added a new table: protein_tree_attr :) It is used to 
store all the tags that are actually present in every node.

BTW, this will change again in e66, as we will merge the protein trees 
and the ncRNA trees in a single set of tables

Regards,
Matthieu

On 10/01/12 16:32, Moretti Sébastien wrote:
> Using your db, I got
> 8 Euteleostomi
> the normal behavior I guess.
>
>
> With our ensembl db, I got
> 8 MISSING
>
>
> We use truncated db to save some space. I guess some tables are missing
> compared to what we used for ensembl 61.
> I will try with a full installation.
>
> Here are tables we load:
> - species core db
> *_core_*.sql.gz
> assembly.*
> coord_system.*
> dna.*
> exon*
> meta*
> seq_region.*
> transcript.*
> transcript_stable_id.*
> translation.*
> translation_stable_id.*
>
> - compara db
> ensembl_compara_*.sql.gz
> genome_db.*
> member.*
> meta.*
> ncbi_taxa_name.*
> ncbi_taxa_node.*
> protein_tree_member.*
> protein_tree_member_score.*
> protein_tree_node.*
> protein_tree_stable_id.*
> protein_tree_tag.*
> sequence.*
> subset_member.*
>
> Thanks
>
>> Hi Sébastien
>>
>> The following code is working for me, and prints the taxon name of the
>> tree starting at node_id=8 (also works with 248).
>>
>> my $tree = $protein_tree_adaptor->fetch_node_by_node_id(8);
>> my $tax_level = $tree->get_tagvalue('taxon_name');
>> print $tree->node_id, " ", (defined $tax_level ? $tax_level :
>> "MISSING"), "\n";
>>
>> Is it working for you as well ? If so, the error must be a side effect
>> of another part of your script. Could you then attach it ? (at least its
>> beginning)
>>
>> Thank you,
>> Matthieu
>>
>> On 10/01/12 15:48, Moretti Sébastien wrote:
>>>> Hi Sébastien
>>>>
>>>> There was indeed some work on tags in rel 65, but this should be
>>>> transparent for users.
>>>> However, missing tags are now returning "undef" instead of an empty
>>>> string. Does it match your error ?
>>>
>>> I have always tested $tax_level this way:
>>> if ( defined $tax_level && $tax_level ne '' ){ ... }
>>>
>>> Currently the script does not go into this if statement.
>>>
>>>> That said, every internal node should have a 'taxon_name' tag, and
>>>> leaves shouldn't (same behaviour as in rel 64). Can you give me an
>>>> example of a failing $node_id ?
>>>
>>> Here are 2 failing node_id
>>> 8 (for ENSGT00470000042273)
>>> 248 (for ENSGT00530000064177)
>>>
>>>
>>> Sébastien
>


-- 
Matthieu Muffato, Ph.D.
Ensembl Developer - Comparative Genomics
European Bioinformatics Institute (EMBL-EBI)
Wellcome Trust Genome Campus, Hinxton
Cambridge, CB10 1SD, United Kingdom