...

One key use of GBIF-merged specimen records is to count or plot the number of organisms in an area. When a wide net is thrown around the globe, the duplicate records are caught and return overstated counts. Ideally a GUID would identify a single unique organism record and enable duplicates to be identified, but I can see no easy way for that to occur within LSID.

There is some discussion of how this might happen with LSID in http://efgblade.cs.umb.edu/twiki/bin/view/BDEI/AbstractEntities

Please feel free to contribute to it.

Because I don't really like to use Wiki's I am going to continue this thread here. There are some useful examples abstract and "real" instantiations of LSID 's at: http://www.i3c.org/wgr/ta/resources/lsid/docs/LSIDqueriesAndResponses.txt

Here is a relevant section:

IsAbstractOf Query: LSIDs may name abstract concepts suchs as "rat myoglobin" as well as concrete instantiations of those concepts such as "rat myoglobin in FASTA format". This query is used to find all of the LSIDs naming concrete instantiations of a particular LSID that refers to an abstract concept. In english, the query string below literaly reads: "Fill in the blank, 1AFT is an abstract concept of _________?" Another way to think about it is, "What concrete instantiations of the abstract concept '1AFT' exist?" The response includes the LSIDs of the 1AFT data in four different formats.

Query string: URN:LSID:pdb.org:PDB:1AFT:URN:LSID:i3c.org:predicates:isAbstractOf:?x

Response: URN:LSID:pdb.org:PDB:1AFT-PDB: URN:LSID:pdb.org:PDB:1AFT-mmCIF: URN:LSID:pdb.org:PDB:1AFT-JPG: URN:LSID:pdb.org:PDB:1AFT-FASTA:

I can think of lots of similar examples in our domain.

Browsing http://www.i3c.org/ and http://www-124.ibm.com/developerworks/oss/lsid/ will generally provide you with examples similar to all of our issues. Somewhere in all those articles there must be examples of how to recognize the identify of a sequence that has been deposited in two different sequence databases. Lots of good reading for the long plane trip to NZ.

Julian

Julian Humphries DigiMorph.Org Geological Sciences University of Texas at Austin Austin, TX 78712 512-471-3275