add swine medical models

[nlwashington]

this page: http://nsrrc.missouri.edu/index.asp

has a nice listing of pig research models. i don't see any way of getting the data that underlies this without scraping. perhaps we can email the maintainers of the website at NSRRC@missouri.edu.

these might need to be hand curated, but there's only a few. let's first check with them to see if they maintain any properly identified data, otherwise we'll curate.

[nlwashington]

i have sent an email to their helpdesk.

[nlwashington]

having reviewed the info on their site, i think we should just hand-curate this resource, and then have some process via dipper that will notify us if there are new models that become available through this site.

@nicolevasilevsky can you help with this task? there are ~80 lines lines outlined on this page: http://nsrrc.missouri.edu/StrainAvail.asp each is annotated to some kind of research area, and a subset are linked to individual pages with references.

the task here would be to iterate through each strain (linked from the identifier), and:

• map the text-based citations to PMIDs or dois
• try to encode the genes and/or variations (using identifiers) present in the animal. this will probably require looking up the genes described and mapping them to NCBIGene ids
• map the "application" to the relevant DOID or Uberpheno term.
• sometimes it might require reading over one of the cited papers to get more accurate info
• save the link for any of the images.

for example, NSRRC:0011:

• under genes, it says, "express a humanized Caenorhabditis elegans gene, fat-1, encoding an n-3 fatty acid desaturase". i read through the cited article, and it turns out that this same transgene was also put into a mouse in a prior publication. i searched for 'fat-1' at MGI, and it's this one: MGI:4354100 based on the description and matched citation. so you could curate this MGI ID as the variant. but because MGI doesn't actually annotate that this allele is primarily the C. elegans gene, can you also add that as the gene? I searched for "fat-1" at wormbase, and got WormBase:WBGene00001393
• the research applications:
  • reproduction ==> HP:0000078
  • cardiovascular ==> HP:0001626
  • immune system ==> HP:0002715
  • depression ==> HP:0000716
  • cancer ==> HP:0002664
• i would map the referenced articles to: PMID:17713802, PMID:16906441, PMID:16672718, PMID:16565727 (i got these by clicking the links for each reference, and snagging the PMID on the pubmed page)

this could be done in a simple excel doc, and exported as tsv. let's put it into one of our monarch gits.

[nlwashington]

@mbrush should these animals just be instances of NCBITaxon:9823 (pig) or is there another class that we should use to indicate an organism (just like for humans we have foaf:Person)?

[nlwashington]

@nicolevasilevsky took a first pass on the data here: https://docs.google.com/spreadsheets/d/1cBdWOs2Hk4z3Bbjvz1jEzj0Oh9eOuMbX9ePtzjydkdk/edit#gid=0

This will need a final pass to make it uniform for processing by dipper. Once cleaned up, it should be exported to TSV and long-term storage in some monarch-data git repo (TBD).