add mp:Reference

[jodischneider]

Add mp:Reference. This should be the OA:hasSource associated with the target of the OA that points to the mp:Data, mp:Material, mp:Methods.

[jodischneider]

mp:Reference, as used in the original paper, is the backing for a claim NOT DISCUSSED in the paper. See Figure 7 page 15 and Figure 8 page 16 of http://www.biomedcentral.com/content/pdf/2041-1480-5-28.pdf

The use case in the paper is mainly digital abstracts -- so that MP1 mp:represents Attribution1

Ways to resolve this:

(a) add mp:represents pointing to an identifier for the paper [this is appropriate if there is only one main paper providing supporting evidence (as opposed to Reference backing) for an MP]

(b) if MP is going to discuss evidence from multiple sources, mp:represents is misleading. Should we mint another relationship? Clarify the scope of mp:represents?

[timclark]

In the example on p15 (fig 7) mp:Reference to Harrison et al. was provided by Spilman et al in support of their claim, made in the paper, that Rapamycin inhibits the mTOR pathway.

The overall claim they are making has three components from a logical point of view, each of which they assert and provide backing references for:

1 - Rapamycin inhibits mTOR pathway 2 - PDAPP J20 transgenics are a (“good” implied here but they go with “popular") mouse model of Alzheimer disease 3 - Feeding Rapamycin to these mice reverses AD symptomatology and pathology

So I’m not quite sure what your comment means, but maybe a skype would clarify?

Tim

On Aug 3, 2015, at 2:34 PM, Jodi Schneider notifications@github.com wrote:

mp:Reference, as used in the original paper, is the backing for a claim NOT DISCUSSED in the paper. See Figure 7 page 15 and Figure 8 page 16 of http://www.biomedcentral.com/content/pdf/2041-1480-5-28.pdf http://www.biomedcentral.com/content/pdf/2041-1480-5-28.pdf The use case in the paper is mainly digital abstracts -- so that MP1 mp:represents Attribution1

Ways to resolve this:

(a) add mp:represents pointing to an identifier for the paper [this is appropriate if there is only one main paper providing supporting evidence (as opposed to Reference backing) for an MP]

(b) if MP is going to discuss evidence from multiple sources, mp:represents is misleading. Should we mint another relationship? Clarify the scope of mp:represents?

— Reply to this email directly or view it on GitHub https://github.com/dbmi-pitt/DIKB-Micropublication/issues/26#issuecomment-127364171.

[jodischneider]

Right. mp:Reference is used for the backing references. But for the original Spilman et al paper you don't use mp:Reference (because it's not referring to itself). This makes sense for a digital abstract -- the MP mp:represents the Spilman et al paper.

In our case, the MP is not specifically intended to be a digital abstract of the document (paper/drug label/...). (It happens to do this to a certain extent.) The main goal, though, is for the MP to represent the claim and all its supporting evidence--which may come from multiple papers that don't a priori cite each other.

So it's not clear that our MP "mp:represents" the paper (it doesn't fully represent it; and there might not be a 1-1 relationship between the MP and the document).

Does that make sense, Tim? If not, let's try to talk later this week.

[timclark]

Not clear to me so let’s talk - I get the distinction in use cases but not your approach

On Aug 3, 2015, at 4:19 PM, Jodi Schneider notifications@github.com wrote:

Right. mp:Reference is used for the backing references. But for the original Spilman et al paper you don't use mp:Reference (because it's not referring to itself). This makes sense for a digital abstract -- the MP mp:represents the Spilman et al paper.

In our case, the MP is not specifically intended to be a digital abstract of the document (paper/drug label/...). (It happens to do this to a certain extent.) The main goal, though, is for the MP to represent the claim and all its supporting evidence--which may come from multiple papers that don't a priori cite each other.

So it's not clear that our MP "mp:represents" the paper (it doesn't fully represent it; and there might not be a 1-1 relationship between the MP and the document).

Does that make sense, Tim? If not, let's try to talk later this week.

— Reply to this email directly or view it on GitHub https://github.com/dbmi-pitt/DIKB-Micropublication/issues/26#issuecomment-127395545.

[jodischneider]

Yifan will implement using the PMID URLs associated with the implementation: http://dx.doi.org/10.1016/S0009-9236(98)90151-5 mp:supports :C1 ; rdf:type mp:Reference .