Storage, preservation, traceability and retrievability of research data
The original/source (unprocessed, raw) data and experimental protocols and the books and digital media where they are stored, and which are originated from research conducted at OSR, are the property of OSR, unless otherwise established though specific contracts with third parties, and may not leave the premises without a legal basis at any time. OSR-sanctioned group leaders are granted management privileges over the project data and may govern their publication and copyright (the latter subject to OSR evaluation). However, they are duty-bound to report any invention to the OSR Office of Biotechnology Transfer in accordance with OSR regulations.
All primary data must be stored, in real-time, on a remotely backed-up research-dedicated central OSR server and/or specifically allocated OSR-provided cloud-based storage accounts, and must remain accessible to the authorized employees for at least five, ideally ten, years after its collection3. It is acknowledged that data generated by analytical instruments (e.g. FACS, PCR, ELISA readers, etc.) might not be immediately backed-up by the instrument hardware and generally needs to be transferred to the personal OSR-provided PC/account of the user, upon which the data becomes available for back-up and storage. All data should be associated with README files providing legends and appropriate indications to allow understanding (for an expert reader) of content and including referral to the appropriate entries in the log books (whether digital or paper-based). Failure to comply with these indications is considered to be at the very least, a DRP but depending on the seriousness and case scenario, may also amount to research misconduct (RM). Investigators are also required to comply with OSR internal policy on the use of IT resources.
It is also a DRP if at any given time the sole repository of primary data is a non OSR computer and/or a computer that is not in the immediate availability of the project leader (and later the corresponding author) during the editorial process or after publication and not accessible by an authorised IT official. In case of a dispute that requires consultation of original data, failure to have complied with the above indications may amount to RM.
For the above reasons and in general to preserve the integrity of the data, upon submission of a scholarly article, book, or PhD dissertation manuscript, throughout the editorial or revision process and after publication, all files containing the underlying data4, should be compiled into a clearly identified directory (folder) for each publication accompanied by appropriate README files; this directory must remain accessible for the legally required length of time, but for no less than five years following the publication of the manuscript. A thoughtful and comprehensive organisation of file/folder structure is key to make it easy to locate and organise data and versioning. Collaborative work, which is often the norm, enhances the need for clear, organised file structure5. Furthermore, data safety and integrity and clear metadata are also relevant for industrial collaborations or in the process of patent protection, particularly in the case of any arising dispute.
In the case of collaborative studies with third parties (e.g. academia, industry, etc.), and if the corresponding author (CA) of the consequent publication(s) is an OSR investigator, it is their duty of care to unsure that an integral copy of the source data underlying experiments conducted by all the collaborating scientists(s) and used in the publications (s) is stored at OSR according to the provisions set forth in this document. Should this not be possible because of the extremely large size of the files (e.g. genomic or imaging data), the collaborating Institution should sign an agreement to make such data readily available upon request from the CA or OSR. In the case of collaborative studies with third parties and if the participating OSR investigator is not the project leader (and later the corresponding author) it is nevertheless the OSR investigator’s duty to ensure that his/her data contributed to the manuscript is collected and stored at OSR as defined above.
Typically, a legitimate request to produce the original (source) data underlying a publication e.g. from a scholarly journal requesting clarifications and/or an investigation body and/or an author, should be addressable within a day or two.
The responsibility for data storage for a research group lies with the respective OSR-sanctioned group leader, who is also responsible for its appropriate and unambiguous logging. It should be noted however, that for all intents and purposes, the scientific community will consider the CA ultimately responsible for the storage, preservation, retrievability and shareability of the original data underlying the publication even if said CA is not the group leader6.
All researchers participating in a given study are permitted to make/store copies of the data if this is compatible with the applicable law and provisions of data protection in case of studies involving humans/clinical data.
Finally, OSR investigators must make exclusive use of their OSR email account for all professional exchanges.
3 Different journals/publishers of funding agencies may have different requirements, to which the CA adheres upon publication of funding. It should be noted that some scholarly journals and funding agencies expect a 10-year accessibility period.
4 With the exception of extremely large source data files (e.g. genomic or imaging data) which may be stored in appropiately safe and backed-up core facility databanks. Such locations should be nevertheless clearly indicated in the README file.
5 There are many available sources of information and tips on research data organisation. One good example is here: ukdataservice.ac.uk/manage-data/format/organising.
6 In such cases where the CA is not also the group leader, a co-corresponding authorship with the OSR-sanctioned group leader is encouraged to better ensure compliance with OSR data ownership guidelines.