Re: [BioMart Users] [biohackathon:270] Text-mining ontological data

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: [BioMart Users] [biohackathon:270] Text-mining ontological data

Arek Kasprzyk
Hi Martin,
this is great. it will be good enough for us.
Joachim: did you have any luck with collecting use cases for this project?

a

On Fri, Aug 26, 2011 at 1:16 AM, Martin Gerner
<[hidden email]> wrote:

> Hi Arek,
>
> I can make sample data available for each of the entity types, if you are
> still interested. It's in an SQL table at the moment, so it'd be formatted
> as a simple tab-separated value file I imagine.
>
> -- Martin
>
> On 26/08/2011 02:16, Arek Kasprzyk wrote:
>>
>> Hi Goran,
>> is there are test dataset available right now that we could use for
>> quick prototyping?
>>
>> a
>>
>>
>> On Thu, Aug 25, 2011 at 1:03 PM, Goran Nenadic
>> <[hidden email]>  wrote:
>>>
>>>>>> As for the data, we are finalising a datastore of mentions of genes,
>>>>>> proteins, species etc. for the entire Medline and (open-access) PMC -
>>>>>> e.g. some 80 million mentions of genes/proteins, associated with
>>>>>> anatomical location where available.
>>>>>
>>>>>   Sounds great! Will that be a continuation of www.pubmed2ensembl.org?
>>>
>>> The aim of pubmed2ensembl is to directly link genomic data (Ensembl) and
>>> biological literature (PubMed) on a large scale. There are links between
>>> over 2 million articles and nearly 150,000 genes from 50 species, and
>>> these
>>> are then integrated with other Ensembl data on these genes for integrated
>>> querying. One can query both PubMed for a particular term (e.g. disease,
>>> process) and get associated genomic data directly, or the other way
>>> around -
>>> e.g. use genomic coordinates as input and get associated papers. The data
>>> is
>>> available for browsing or download at www.pubmed2ensembl.org, and the
>>> associated paper was accepted two days ago (PLOS ONE), and is in
>>> production
>>> now :).
>>>
>>> The new dataset that Martin mentioned will be much larger, including
>>> links
>>> from some 200 million mentions of various 'things' in both PubMed and
>>> PubMed
>>> Central - from genes, species, anatomical terms etc., many of them linked
>>> to
>>> relevant databases and processes they are involved in. As I said, this
>>> dataset will be available very soon and it could be also potentially
>>> integrated in a biomart. I guess with all Joachim's efforts one would be
>>> also able to use SPARQL to query this data (if or once it's available via
>>> BioMart 0.8).
>>>
>>> Best,
>>> Goran
>>>
>>>
>>>
>
>
>
_______________________________________________
Users mailing list
[hidden email]
https://lists.biomart.org/mailman/listinfo/users
Reply | Threaded
Open this post in threaded view
|

Re: [BioMart Users] [biohackathon:270] Text-mining ontological data

Joachim Baran-2
Hi Arek!

> Joachim: did you have any luck with collecting use cases for this project?
  The short BioMart tutorial/demo that I gave was mainly intended for the text mining group, even though it was quite a general demonstration of BioMart's data integration facilities.

  I think the new impressions about BioMart need to sink in first before the text miners can come forward with desired features that would need supporting. Currently, we are already supporting to link gene, species or other normalisations to any of the existing 35 databases in the BioMart Central Portal and all other large or small mart instances.

Joachim
_______________________________________________
Users mailing list
[hidden email]
https://lists.biomart.org/mailman/listinfo/users
Reply | Threaded
Open this post in threaded view
|

Re: [BioMart Users] [biohackathon:270] Text-mining ontological data

Arek Kasprzyk
ok, i can come up with something. the text mining use cases are
independent from BioMart implementation.
We only need general ones. We'll be able to design the system around them.

thanks
a

On Sat, Aug 27, 2011 at 2:08 AM, Joachim Baran <[hidden email]> wrote:
> Hi Arek!
>
>> Joachim: did you have any luck with collecting use cases for this project?
>  The short BioMart tutorial/demo that I gave was mainly intended for the text mining group, even though it was quite a general demonstration of BioMart's data integration facilities.
>
>  I think the new impressions about BioMart need to sink in first before the text miners can come forward with desired features that would need supporting. Currently, we are already supporting to link gene, species or other normalisations to any of the existing 35 databases in the BioMart Central Portal and all other large or small mart instances.
>
> Joachim
_______________________________________________
Users mailing list
[hidden email]
https://lists.biomart.org/mailman/listinfo/users