Unique and stable IDs for samples, datasets, …
For centuries, scholars and librarians alike have been accustomed to using references to books, journals, and articles. Those references are unique and stable, and while their exact format may differ depending on the actual context, they always contain all necessary information to uniquely identify the ressource. In the 20th century, this has been accompanied by numbers such as ISBN and ISSN, and lately digital object identifiers (DOIs) in the digital age.
While there is a broad consent how to reference all kind of printed ressources, nothing like that exists for other types of objects scientists deal with routinely in their research, such as samples, datasets, and alike.
Recently, extending the concept of the DOI to datasets accompanying research articles has been brought forward. While DOIs depend on an external entity, one can use a similar scheme within the own lab to uniquely identify everything that needs to be referenced, be it a sample, a dataset, a “recipe” describing processing and/or analysis of data, versions of software, and alike.