Persistent identifiers and associated metadata describe resources such as datasets, software, publications, people, research organizations, funders, and grants. An important part of this metadata is the description of connections between these resources. Together these resources and their connections form a graph, the PID Graph (Fenner & Aryani, 2019).
Accessing information available in this PID Graph, while preserving the rich connections between resources, is not trivial, and the JSON REST APIs that most PID service providers are providing to users, while having a very good track record allowing users to access a single resource or a list of similar resources, might not be the best fit for more complex queries of the PID Graph.
Enter GraphQL, a query language that uses a graph as the underlying data model and aligns well with the kinds of queries that need to be supported in the PID Graph. GraphQL was started by Facebook in 2012, made available as Open Source software in 2015, and in 2019 has become a mainstream technology with broad support in terms of libraries, tools and services. In May 2019 DataCite launched a pre-release version of a GraphQL API to query for PIDs and metadata from scholarly resources registered with DataCite, Crossref, ORCID, [ROR](https://ror.org], and others. The production version of this API will launch in April 2020.
A browser based GraphQL client is available at https://api.datacite.org/graphql, and documentation can be found at https://support.datacite.org/docs/datacite-graphql-api-guide. Use the PID Graph category of the PID Forum to post questions, comments and announcements.