Sir

We support the suggestion made by Carlos Santos and colleagues in Correspondence (Nature 438, 738; 200510.1038/438738a) that data associated with peer-reviewed articles should be submitted to recognized, public repositories wherever possible.

We suggest that attaining this goal requires the support of national and international funding bodies that are willing both to implement data policies and to fund efforts to create community-driven standards and public repositories.

For funding bodies, supporting a data policy can be expensive. Current policy for the UK Natural Environment Research Council (NERC; http://www.nerc.ac.uk) states that all data generated by projects funded under the environmental-genomics programme and the post-genomics and proteomics programme must be submitted to a suitable public repository, when one is available. NERC puts approximately 12% of the funds from each of these programmes towards data management and training through the establishment of the NERC Environmental Bioinformatics Centre (http://nebc.nox.ac.uk), which facilitates ‘omic’ data management by developing data standards, software, databases, bioinformatics workstations and courses, and delivering these to the community, as well as hosting digital data for cases where suitable public repositories do not already exist.

We argue that putting the tools and facilities in place to enable good data management is an area worth investing in. As well as addressing the aims of integrated, long-term data storage and access, this investment would minimize duplication of effort, facilitate uptake and sharing of data and maximize the potential for comparative analyses.