MarathonMouseUseCase

From W3C Wiki

The overall story: The PPARD agonist GW501516 is in Phase II trials for dyslipidemias. It was recently shown to have a remarkable effect on mice, making them able to run twice as far as controls, and making them resistant to the obesigenic effects of high fat diets. This is of course, one of the great outstanding markets left for blockbuster drugs: Imagine, the benefits of exercise in a pill! Of course, no drug is without problems, and PPAR-related compounds generally have some scarey toxicity problems, particularly around cancer.

This story can be told from several perspectives:

  • Business: who has a financial interest in the compound, what are the issues with respect to its path toward FDA approval, what are the competitors.
  • Genetics: information about the target, PPARD. The various human subtypes and polymorphisms. Relationship between human PPARD and mouse (where most of the studies have been done). Role in normal biology (in addition to its fat-related effects, it plays a role in development, inflamation and possibly neurological activity).
  • Toxicity: PPARD agonists have increased growth rates of certain kinds of cancer cell lines, and the FDA has special guidelines for showing that PPAR-related compounds are not carcinogenic.

1 Basic information about the gene and the drug(s)

  • Formal gene name: peroxisome proliferative activated receptor, delta Symbol: PPARD NB: there are naming problems -- generally PPAR beta is the same same as delta. Also note there are two human isoforms (although not two mouse isoforms).
  • The ppard agonists (potential drugs) are currently called GW0742 and GW501516. These are very closely related compounds (just a single atom difference between them). Here is an article that
describes their structures. It would be really nice to be able to extract these structures from the article and get them in RDF form.

2 Literature searches of pubmed for (ppard or "ppar delta") yields 77 articles. A search on GW501516 gives 11 articles. The key articles are:

 2.1 The "Marathon Mouse" article itself (including videos of the mice). 98 related articles as found by the Pubmed "related articles" link.
 2.2 The initial publication of the agonist GW501516, on the cholesterol indication. The 107 related articles from Pubmed "related articles".  Question: what is the overlap between the related articles here and the related articles in 2.1, and which of those (if any) do not appear in the original searches on PPARd and GW501516?
 2.3 A lipid response study of GW501516 using Affy arrays.  The raw data is here.  NB: this expression array data is not submitted to a public repository, but is served off the LSBM web site.  Table 1 of the supporting information for that publication is also pretty interesting.  Could it be extracted from the PDF and represented in RDF?
 2.4  Two articles (one two) about PPARD's role in obesity.
 2.5 Its role in inflamation.
 2.6 A hint of a role of PPARd (called here PPARb) in neurological disorders
 2.7 A big issue is whether PPARd plays a role in causing or promoting cancer or has other toxicities.  Here are four articles on various aspects of that question: a nice summary two three four
 2.8 Here are two articles about PPARd related drugs and how to synthesize them.  The data in the tables of the first would be very valuable.  the second is also interesting. 

3. Databases to integrate. There is information about PPARd in many publically available databases. Most of this data is linked to from NCBI's Entrez Gene entry (see 3.1). Data sets not accessible through links from NCBI are highlighed.

3.1  TheEntrez Gene entry and the UniProt entry
3.2 Refseq entries isoform 1 nucleotide isoform 1 protein isoform 2 nucleotide isoform 2 protein
3.3 The homologene report
3.4 The NCBI polymorphisms.  I would like to find other polymorphisms, since these differences should explain the naturally occurring differences among people.   Here is the NIEHS snp data.  And here is some more PPARd snp data from Utah's Gene SNPs project.  And now four mutations automatically extracted from the literature by the NucleaRDB project. 
3.5 The GeneCards entry summarizes data from other databases.  It's particularly interesting for the tissue distribution data.
3.6 Not only is the structure of the protein known, but we have a structure of it bound to its ligand!
3.7 Here is the OMIM entry which both provides a nice summary of what is known about the gene and a description of its known role in various diseases.
3.8 Here is the summary page for records in the GEO database of expression array data relevant to PPARd.   Some datasets of particular interest include three on fat metabolism: one two three.
3.9 Here is the UCSC Genome Browser entry  Many tracks (and corresponding aspects of the underlying database) are of potential interest here.  For example,  regions upstream of the gene that is conserved between mouse and human (a track) would be a particularly interesting place to look for human polymorphisms (another track) because those might suggest regulatory polymorphisms among individuals that would explain individual human differences.
3.10 There isn't very much pathway data that I could find. KEGG shows it is a downstream responder in the WNT pathway.  The CGAP entry has pointers to better pathway data from Biocarta: ppar pathways and lipid metabolism / toxicity pathways.
3.11 The signaling gateway entry doesn't offer pathways, but does have links to bind protein-protein interactions.
3.12 Here is the PPARd information from the MGI database (in html format).  Here is general information about access through the  MGI SQL server.  Ian sent around access passwords, etc. via email.  

4. Business links (how to find these? Google?)

4.1  The compound is owned by Ligand and licensed to GSK. Among many other places, financial information about the company ( nasdaq:LGND ) can be found here.  I imagine there are better sources of RDF business information out there.
4.2 The FDA is particularly concerned about safety of PPAR-related compounds, and has published specific safety standards just for PPAR compounds.