Submission regarding the Draft NIH Genomic Data Sharing Policy
19 November, 2013
Office of Science Policy
National Institutes of Health
6705 Rockledge Drive, Suite 750
Bethesda, MD 20892
Submission regarding the Draft NIH Genomic Data Sharing Policy
1. Background to Submission
The Human Variome Project is an international consortium of scientists and healthcare professionals who are working towards a significant reduction in the burden of genetic disease on the world’s populations. The aim of the Human Variome Project is to ensure that all information on genetic variation can be collected, curated, interpreted and shared freely and openly. This will lead to speedier, better and cheaper diagnosis and treatment of genetic disorders, and better insight into the causes, severity and effect of common disease. The Human Variome Project achieves its aims by establishing and maintaining the necessary standards, systems and infrastructure, by providing education and training to scientists, clinicians, genetic counsellors, other healthcare professionals and the general public and by assisting nations build their capacity in medical genetics and genomics. The Human Variome Project acts as an umbrella organisation and works to encourage communication and collaboration around its central vision.
The importance of the Human Variome Project was recognised in 2011 by the United Nations Educational, Scientific and Cultural Organisation in the Project’s admittance to Consultative Partner status.
2. General Comments on the Draft Policy
We are very pleased to see that the NIH is taking an active interest in data sharing and will be, through this policy, mandating the submission of data generated by the array-based and high-throughput research projects they fund. The Human Variome Project exists to facilitate the free and open sharing of just this type of information and we applaud the NIH on taking the first steps down the road to full data sharing.
However, we feel that this policy does not yet go far enough in terms of scope or vision, to fully enable and promote the routine sharing of genetic variation information by activities funded by the NIH. We have outlined our concerns in detail in the sections below.
3. Comments on Specific Sections of the Draft Policy
3.1 Scope and Applicability (Section II)
The Policy as drafted makes mandatory the submission of data generated by “all NIH-funded research that involves large-scale human and nonhuman genomic data produced by array-based or high-throughput genomic technologies, such as GWAS, SNP, whole-genome, transcriptomic, epigenomic, and gene expression data, irrespective of funding level and funding mechanism.”
This is an admirable start, but we must ask why limit this policy to “large-scale” genomic data? There is undoubtedly much data generated by NIH-funded research projects that would not be considered “large-scale” that has potential value, both in terms of re-use in future research projects and also clinically, when diagnosing and treating patients afflicted with rare, genetic disorders. The technical and administrative effort required to submit such data, to an NIH-funded repository (or otherwise) is not burdensome and could be easily integrated into existing research reporting workflows.
The NIH, in drafting this Policy, obviously acknowledges the benefits of wide-scale data sharing. It seems unnecessarily restrictive and potentially confusing, to limit the scope of this policy to certain research projects that are categorized only under the ambiguous phrase “involves large-scale…genomic data.”
3.2 Human Genomic Data (Section IV.C)
We are pleased to see that the draft policy does not seek to restrict the submission of data by research projects within the scope of the Policy solely to NIH-funded repositories. Although NIH-funded repositories play a key role in the emerging data-sharing infrastructure that is operating globally, they are by no means the only places that such data could be stored and shared. Indeed, in quite a large number of cases, future research projects and clinical practice would be better served if the data from NIH-funded research projects were shared much more broadly than between NIH-funded repositories. Numerous disease-specific research and clinical communities are well served, and have been well served for over a decade, by international gene/disease specific databases (GDSDBs) or locus specific databases (LSDBs). While it could be argued that individual researchers working in these disease areas would be aware of these extant resources and include submission to them in their data sharing plans, and this might very well be the case. However, we still believe that it would be wise for the NIH to include a more direct acknowledgement of data sharing options beyond those funded by the NIH. The desired end result of this Policy is obviously to better human health through improved research and clinical delivery: encouraging researchers to think more carefully about where the submission of their data would be most useful would go a long way to reaching this end result.
Finally, we are a little concerned over the implied insistence of submission to an NIH-funded repository as the minimum requirement of this policy, as indicated by the sentence: “NIH-designated data repositories need not be the exclusive source for facilitating the sharing of genomic data.” Our concerns are related to how this policy would be applied to those research projects that generate data, via overseas partners and collaborators, from samples derived from non-US citizens and within a different ethical, legislative or social context. In these situations, researchers may have difficulty, or be prohibited from sharing data, either completely, or in part, to a repository that is operated within a different country and funded by a separate sovereign power. While we are sure that such a circumstance would be taken into consideration if articulated in the required data sharing plans, the draft Policy itself is silent on the matter and would benefit from this matter being addressed.
4. Concluding Remarks
The Human Variome Project fully endorses the proposed draft Policy. While we believe that the Policy could be extended to further promote and encourage the free and open sharing of data to enable more expansive use within both research and clinical contexts, it is a good first step.
Richard G.H. Cotton AM PhD DSc FRCPA (Hon.)
Human Variome Project
for the Human Variome Project International Scientific Advisory Committee