Re: [Treebase-devel] Rod Page on TreeBASEII

[William P. <wil...@ya...>]

On May 25, 2010, at 8:41 AM, Hilmar Lapp wrote:

> Instead, we should construct a new LinkOut mapping file for NCBI and submit it to their ftp server at the appropriate location. That would be the clean and Right Way.

Ah... I didn't know the outlinks could be submitted that way. I was preparing the send off the (below) message to Scott Federhen. 

If this letter to him is still relevant, please make comments / edits.  I'm thinking that I should remove the bit about returning a list of studies / matrices / trees because those links don't seem to work -- they are returning the taxon tab instead of jumping to the respective study / matrix/ tree tab. 

Bill



=====================================================
On May 1, 2003, at 6:21 AM, Scott Federhen wrote:

> Hi Bill:
> 
> The TreeBase links went up in Entrez yesterday -
> you can find all of them with:
> 
>  http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=Taxonomy
>  search for "loprovtreebase[filter]" (no quotes) and hit 'Go'
>  switch from 'Summary' to 'LinkOut' and hit 'Display'


Dear Scott,

Thanks very much for posting NCBI's outlinks to TreeBASE over the last seven years. 

TreeBASE has now been redeveloped from the ground up. Among the new features, the following items are relevant to NCBI's outlinks:

- TreeBASE has a new RESTful API based on the PhyloWS standard

- Regrettably, TreeBASE has had to reissue new taxon id numbers. The old version was essentially just a distinct list of taxon labels. The new version is the result of a mapping between taxon labels in each study and a taxonomy, in which homonyms receive distinct IDs. Since our concept of what is a "taxon" has changed (from unique string to a vetted list) the ID numbers for these has also changed.

- TreeBASE now tries to store the equivalent NCBi taxid for each taxon in TreeBASE. This means that Entrez can outlink to TreeBASE by using NCBI taxids rather than having to store any of TreeBASE's IDs. All that's needed is for NCBI to know which of its taxa are also in TreeBASE. 

I am attaching a list of all NCBI taxids that current map to data objects in TreeBASE. The PhyloWS format for creating an outlink to Homo sapiens in TreeBASE (using NCBI taxid 9606) is as follows:

http://purl.org/phylo/treebase/phylows/taxon/find?query=tb.identifier.ncbi=9606&format=html

or if it's safer to use %3D as a substitute for the equals sign:

http://purl.org/phylo/treebase/phylows/taxon/find?query=tb.identifier.ncbi%3D9606&format=html

If you want to return a list of studies, use the following:

http://purl.org/phylo/treebase/phylows/taxon/find?query=tb.identifier.ncbi%3D9606&format=html&recordSchema=study

If you want to return a list of matrices (i.e. alignments), use the following:

http://purl.org/phylo/treebase/phylows/taxon/find?query=tb.identifier.ncbi%3D9606&format=html&recordSchema=matrix

If you want to return a list of trees, use the following:

http://purl.org/phylo/treebase/phylows/taxon/find?query=tb.identifier.ncbi%3D9606&format=html&recordSchema=tree


I hope you're able to update your outlinks without too much trouble. I'd be happy to provide these data in a difference format, if it's helpful. 

regards,

Bill