INTRODUCTION In the current era of large-scale biomedical research, generating and sharing datasets in an open manner is an important, but non- trivial task. However, sharing published data is often low on a researchers priority list for a number of reasons: e.g. journals place limits due to lack of space; making data accessible can be time-consuming/tedious and often this e fort will not be formally recognised. The lack of accessible scienti ic research data is increasingly of concern, not just to researchers, but also to funders, governments and patients 1-2 . Inaccessible data promotes wastage in funding 1 . Lack of publication of ‘non-groundbreaking’ but still valid research can promote bias with serious implications for healthcare 3 . On the other hand, open and accessible data can be beneic ial to scienti ic progress in several ways; for example enabling data to be veri ied 4 or the testing of novel hypotheses that were unforeseen at the time of data generation 5 . F1000Research is working with funders and institutions to begin addressing some of the challenges and promote the publication of research data in an open and accessible way. Helping biomedical researchers gain the credit they deserve Varsha K. Khodiyar a , Cesar A. Berrios-Otero b , Rebecca N. Lawrence a a F1000Research, Faculty of 1000, Middlesex House, 34–42 Cleveland St, London W1T 4LB, UK b F1000Research, Faculty of 1000, 121 West 27th Street, Suite 604, New York, NY 10001, USA F1000Research.com | research@f1000.com | @f1000research | facebook.com/F1000Research ALL STUDY DATA IS ACCESSIBLE All F1000Research articles include the underlying data. We use igshare to host data, and they provide a ‘widget’ within the article which displays the data. Figure 1 shows a typical igshare widget from an F1000Research paper 6 . The widget records altmetrics such as number of downloads and sharing on social media. The dataset has its own DOI and can be cited independently from the paper 7 . The data citation also includes the date of access, so as to facilitate study replication. We are now developing an in-article data plotting tool to enable quick basic manipulation of raw spreadsheet data on-the-ly by referees and readers. REFERENCES 1. Chan A-W et al., 2014 Increasing value and reducing waste: addressing inaccessible research. Lancet 383(9913):257-266 2. Foreign & Commonwealth Oice, 2013. G8 Science Ministers Statement, London 12 June (Accessed: 20 Jan 2014) 3. Goldacre B, 2013. Why – and how – I wrote Bad Pharma Bad Science [blog] 8th October [Accessed 20 Jan 2014] 4. Simonsohn U, 2013. “Just Posting It” works, leads to new retraction in Psychology. Data Colada [blog] 17th September [Accessed: 20 Jan 2014] 5. Chappell, PR and Lorrey, AM, 2013 Identifying New Zealand, Southeast Australia, and Southwest Pacific historical weather data sources using Ian Nicholson’s Log of Logs. Geosci Data J (http://dx.doi.org/10.1002/ gdj3.1) [Early View (Online Version of Record published before inclusion in an issue)] 6. Glusman G et al., 2012. Low budget analysis of Direct-To-Consumer genomic testing familial data [v1; ref status: indexed, http://f1000r.es/ Nmz0oW] F1000Res 1:3 7. Glusman G et al., 2012. Son exome files. figshare. [Accessed 20 Jan 2014] UPCOMING INITIATIVES We work on several initiatives aimed at reducing technological barriers to data sharing. We are working with institutions to provide researchers with a quick and easy submission process for submitting data papers directly from institutional repositories. We are also involved in several projects as part of the Research Data Alliance (www.rd-alliance.org) to look at data work lows within artic les, and bi-directional cross-linking of articles with datasets in repositories. F1000Research is additionally involved in projects aiming to extend the types of scienti ic output that can be recognised for career progression purposes. For example, we are working with relevant stakeholders to discuss new metrics that can be used to assess data generation, data publication and data sharing, as formal contributions to a scientist’s overall impact. TRANSPARENT PEER REVIEW F1000Research is known for being the irst l ife science journal to practice invited post-publication and transparent peer review. Figure 2 illustrates how our peer review process difers from traditional peer review. Transparent peer reviews allow the provenance of each article to be followed, as well as enabling reviewers to claim credit for their work. In collaboration with others, we are working to establish best practice for transparent peer review of the datasets and software accompanying published articles. DATA ARTICLES We ofer authors the option to publish data-only papers, which present the data alongside a detailed description of the protocol used to generate it. Data articles enable researchers involved in the non-trivial task of generating the dataset (and making it accessible) to gain priority and credit for their work. FIGURE 1 Screenshot of a typical figshare data widget with open and accessible raw research data FIGURE 2 Comparison of traditional peer review with F1000Research’s post-publication peer review Article submitted Closed referee process Article Published May have to go around this cycle multiple times. TRADITIONAL JOURNAL Author revises article Data submitted Data in supplemental iles Article submitted In-house pre- refereeing check Author revises article INDEXED Article published Open referee process Data submitted Data in appropriate repository Submission to publication: MONTHS Submission to publication: DAYS . F1000 Posters: mons License. F1000 Posters: Use Permitted Creative Commons License. F1000 Posters: Use Permitted under Creat mitted under Creative Commons License. F1000 Posters: Use Permitted unde ed under Creative Commons License. F1000 Posters: Use P ative Commons License. F1000 P ns Lice