Originally omnicorp stored PMIDs as ints (i.e. just the int part of a PMID:12345 identifier). In moving to allow a broader set of publication identifiers (such as DOI or CORD19 identfiers), we've changed the postgres and redis cache to keep strings of the full curie. The UI needs to understand that change so that it can serve the correct papers for an edge.
Originally omnicorp stored PMIDs as ints (i.e. just the int part of a PMID:12345 identifier). In moving to allow a broader set of publication identifiers (such as DOI or CORD19 identfiers), we've changed the postgres and redis cache to keep strings of the full curie. The UI needs to understand that change so that it can serve the correct papers for an edge.