Translational Vision Science and Technology (TVST) is establishing a Data Science section that will publish Data Science descriptor articles featuring external, scientifically valuable data sets and software libraries relevant to all aspects of vision science.The data sets can vary in nature and may include observational data sets (e.g., from data developed in laboratory experiments such as genome-wide association studies or from registration clinical trials) and computational data sets. Required to be open-access and open-source, these data sets and software libraries are intended for re-use by the scientific community with the goal of accelerating scientific discovery.
The data sets and libraries will incorporate documentation and narrative content with curated, structured descriptions (metadata) of the data set and/or code. The procedures used to develop the data sets will be described in detail in the publication. These descriptions should include machine readable metadata files and must provide information that:
- will enable other investigators to interpret, re-use, and re-analyze the primary data set;
- will enable other investigators to link to the data repository (e.g., figshare or Dryad, or other digital repositories) in which the data are stored (TVST will not host the data set); and
- will enable investigators who have developed the data set to demonstrate to funding agencies that they have fulfilled data-sharing requirements.
Hypothesis testing or extensive analysis of the data set should be provided as a separate publication (preferably in the same issue of TVST) and not in the publication featured in the Data Science section. Data sets and software papers can appear as standalone publications in the Data Science section of TVST, i.e., can be published without an accompanying manuscript that uses the data set or software.
The Data Science publications are subject to peer-review in order to validate:
- the quality of the procedures used to generate the data set or software;
- the completeness/appropriateness of the descriptors;
- the functionality of the proposed software; and
- the re-use value of the data set or software.
These publications are citable and must be cited by investigators who use the data set or software, which will afford the team who created these valuable resources appropriate credit. As with all TVST publications, content in the Data Science section is indexed by PubMed, Scopus, MEDLINE, Google Scholar and Clarivate.
Full release of the research data and/or software upon publication in TVST is mandatory. All content in the Data Science section of TVST is published under a Creative Commons Attribution 4.0 International License (CC BY) to enable maximum re-use of these open access materials. This license allows users to share and adapt the data set/software (including for commercial purposes) provided that publication in TVST is cited as specified by the authors.
The Data Science publication costs are competitive. To promote this important initiative, we are starting with a charge of $300 per article. (Please note that publication costs for the manuscript that reports extensive analyses or hypothesis testing of the data set/library are the routine full TVST publication costs [currently, $1500 for ARVO members and $1850 for non-members].) Authors who cannot afford publication costs can apply to the ARVO Foundation for financial support. Although datasets and software libraries in TVST will be published under the CC-BY license, TVST does not require that the authors use CC-BY for the data or software, just that the authors must use some open-source license.