Sizing the Problem of Improving Discovery and Access to NIH-Funded Data: A Preliminary Study

Abstract

ObjectiveThis study informs efforts to improve the discoverability of and access to biomedical datasets by providing a preliminary estimate of the number and type of datasets generated annually by research funded by the U.S. National Institutes of Health (NIH). It focuses on those datasets that are “invisible” or not deposited in a known repository.MethodsWe analyzed NIH-funded journal articles that were published in 2011, cited in PubMed and deposited in PubMed Central (PMC) to identify those that indicate data were submitted to a known repository. After excluding those articles, we analyzed a random sample of the remaining articles to estimate how many and what types of invisible datasets were used in each article.ResultsAbout 12% of the articles explicitly mention deposition of datasets in recognized repositories, leaving 88% that are invisible datasets. Among articles with invisible datasets, we found an average of 2.9 to 3.4 datasets, suggesting there were approximately 200,000 to 235,000 invisible datasets generated from NIH-funded research published in 2011. Approximately 87% of the invisible datasets consist of data newly collected for the research reported; 13% reflect reuse of existing data. More than 50% of the datasets were derived from live human or non-human animal subjects.ConclusionIn addition to providing a rough estimate of the total number of datasets produced per year by NIH-funded researchers, this study identifies additional issues that must be addressed to improve the discoverability of and access to biomedical research data: the definition of a “dataset,” determination of which (if any) data are valuable for archiving and preservation, and better methods for estimating the number of datasets of interest. Lack of consensus amongst annotators about the number of datasets in a given article reinforces the need for a principled way of thinking about how to identify and characterize biomedical datasets.

Influenza Research Database: An integrated bioinformatics resource for influenza virus research

“…Furthermore, Influenza and Other Respiratory Viruses also reported that the IRD database paper (5) was the #1 cited paper in the journal in 2014. In a survey of research articles that were published in 2011 and had received NIH funding, IRD was the 6th most frequently acknowledged repository for molecular data, behind only some of the major databases managed by the U.S. National Library of Medicine (6). In addition, the IRD website has been heavily used by researchers worldwide, with over 1300 usage sessions per week on average in 2015 as per Google Analytics.…”

FlyClockbase: Importance of Biological Model Curation for Analyzing Variability in the Circadian Clock of Drosophila melanogaster by Integrating Time Series from 25 Years of Research

“…This discussion is relevant because the increasingly data-centric nature of biological discovery has resulted in calls for improved access to existing data (NIH et al . 2012; R ead et al . 2015; W ilkinson et al .…”

“…Thus, our more modest aim here is to highlight lessons we learned while constructing FlyClockbase, hoping they will be useful as the biological community works towards finding more general solutions (NIH et al . 2012; R ead et al . 2015; W ilkinson et al .…”

“…However, it is impossible to address the timely and relevant topic of organizing biological data without the concrete context of a specific resource like FlyClockbase. This discussion is relevant because the increasingly data-centric nature of biological discovery has resulted in calls for improved access to existing data (NIH et al 2012; R ead et al 2015; W ilkinson et al 2016), which is easier said than done (G oldston 2008; D oan et al 2012; G itelman 2013; H uang and G ottardo 2013; M c C allum 2013). As physical access to data is increasing, the next frontier is defined by the ability to efficiently identify datasets of relevance for a given topic.…”

Estimating the scale of biomedical data generation using text mining

“…Since most biomedical datasets are not deposited in a particular biomedical repository rather contained in the article itself(13), how to identify whether biomedical research articles are producing certain kinds of data is not a trivial issue. Like the biological processes and mechanisms being studied, the resulting data is diverse and the terminology describing the data reflects this diversity.…”