Add mappings to phenotypic series in OMIM

[cmungall]

Currently with DO to OMIM xrefs it isn't clear what should be subclass and what should be equivalent

I suggest we add equivalence axioms to PS IDs where they exist
E.g. http://www.omim.org/phenotypicSeries/PS209850

E.g. xref: OMIM:PS209850

We then no longer need to manually maintain xrefs to all the individual sub-types - can be extracted from the up-to-date OMIM ID <-> PS ID provided by OMIM

[cmungall]

First task is to bring in the PS to OMIM mapping into the pipeline, assigning @nlwashington, handed off to me after that

[lschriml]

Good idea.
How will be handle cases where we don't consider some member(s) of the
phenotypic series as equivalent to a DO disease ?

Cheers,
Lynn

On Mon, Feb 9, 2015 at 5:58 PM, Chris Mungall notifications@github.com
wrote:

Currently with DO to OMIM xrefs it isn't clear what should be subclass and
what should be equivalent

I suggest we add equivalence axioms to PS IDs where they exist
E.g. http://www.omim.org/phenotypicSeries/PS209850

E.g. xref: OMIM:PS209850

We then no longer need to manually maintain xrefs to all the individual
sub-types - can be extracted from the up-to-date OMIM ID <-> PS ID provided
by OMIM

—
Reply to this email directly or view it on GitHub
#25.

[cmungall]

unassigning @nlwashington we now have the PS IDs http://nif-crawler.neuinfo.org/monarch/ttl/omim.ttl

@lschriml we can't have anything non-equivalent as an equivalence axiom. We can use an annotation axiom such as dbxref for these if really useful. But let's do an analysis first. I would be curious to see the ones that don't map and the reason why. I would hope we could all come to an agreement on what are sensible groupings.

[cmungall]

What is curious is the cases where no PS exists where we would expect one; e.g. Alzheimers.