The joint principles on data citation represent a new phase of activity that focuses on principled integration with the scholarly research and publishing ecosystem and a broad consensus on data citation practices. What has emerged in the publishing and research communities is agreement that citation is needed to support attribution and verification, recognition that citations must support both human and machine clients, maturity of robust persistent identifiers, and the desire to integrate data citation in standardized ways within publications, catalogs, tool chains and larger systems of attribution. [...] Lowering the barrier to research data discovery and use, coupled with an increased ability to link data with publications, could enable new forms of scholarly publishing, promote interdisciplinary research, strengthen the linkage between policy and science and lower the costs of replicating and extending previous research. For this reason, the submission requirements for Science – one of the most cited, read and respected journals in the sciences – stipulate that “all data necessary to understand, assess and extend the conclusions of the manuscript must be available to any reader of Science” and that “citations to unpublished data [emphasis added] and personal communications cannot be used to support claims in a published paper”..