On February 11-12, 2020, the National Library of Medicine (NLM) held a very interesting workshop on the role of what they are terming “generalist” data repositories (as opposed to discipline or sub-discipline specific repositories, possibly further limited to specific types of data sets). I was able to catch only a small part of the workshop, but what I saw was really excellent. NLM has recently issued a summary of the meeting, and the video recordings of the presentations are also available (there are links in the summary). See
https://datascience.nih.gov/data-ecosystem/NIH-data-repository-workshop-summary
This report is particularly salient in conjunction with the work of the recent National Academies Committee on Forecasting Biomedical Data Life Cycle Costs, which recently issued its final report. This report in essence discusses the role of generalist (in the report’s language “State 3) vs specialized (“State 2”) repositories at great length. I note that Dr. Maryann Martone, who co-chaired the February workshop, served on the National Academies Committee; in the interests of full disclosure, I note I was also a member of this committee. For more about the work of the committee, see
The report is online here:
I would also direct attention to the recent guest blog post by Dr. Susan Gregurick, Associate Director for Data Science and Director, Office of Data Science Strategy, National Institutes of Health, on issues related to generalist repositories at
Finally, let me say that while I think that the biomedical community’s work in this area is particularly advanced and thus a great source of potential insights, these issues are important in all fields of data-intensive scholarship, not just biomedicine.
Clifford Lynch
Director, CNI