Open Science @ FAPESP

Português

Since the beginning of the 90’s, FAPESP has been practicing and fostering initiatives towards Open Science, as a means of facilitating scientific collaboration and production and dissemination of knowledge – including not only scientists, but all those who contribute to or profit from such knowledge, such as policy makers, practitioners, innovators and society as a whole.

The Foundation aims thereby to create conditions for increased scientific, economic and social impact of the research it funds.

Its initiatives towards Open Science are based on the principle that outputs of the research financed by the Foundation are a public good and must be made public as soon as possible, while respecting the principles of scientific ethics, privacy and security, as well as protection of intellectual property.

1. Open Science

Though there is no fixed definition of the term "Open Science", it is usually used to denote the set of policies and actions to disseminate knowledge, usually through digital means, so that all outputs associated with scientific research become accessible to all, reproducible and reusable. These outputs include publications, data, computational processes, software and methodologies used to conduct the research. The ultimate goal of Open Science is to promote innovation and advancement of knowledge through collaboration among scientists and reuse of outputs, thereby accelerating scientific, technological, economic, and social progress.

Digital dissemination of scientific research is usually implemented through publication, in public digital repositories, of sets of files that contain the three basic components of Open Science – Open Access (scientific papers), Open Data (including, among others, raw data, specifications, models, and documentation of experiments), and Open Computational Processes (such as software and algorithms). 

2. FAPESP initiatives towards Open Science

For more than 30 years, FAPESP has been defining policies and practicing initiatives associated with Open Science - through norms, programs, and actions in many fields. Similar to what is practiced in other countries, it aims thereby to enhance the visibility of Brazilian research, fostering its dissemination and thus increasing its scientific, social and economic impact. 

2.1. SciELO

The Scientific Electronic Library Online (SciELO) was created within FAPESP’s grant modality to support research communication. It is a digital collection of open access journals, edited by national scientific societies, professional associations, universities and research institutions.

SciELO’s Advisory Committee periodically advises on which new journals may be invited to become part of its collection, and which journals no longer satisfy its goals in terms of quality of content and regularity.

SciELO was publicly launched in 1998 as a pilot project, in collaboration with BIREME/PAHO/WHO, to create Brazilian know-how in digital publication, and in metrics for evaluation of online scientific journals. A pioneer effort in Open Access, it has since become recognized as one of the most important open access programs in the world for scientific communication. Its publication model has been adopted by 15 other countries, which compose the SciELO Network of collections of open access periodicals that are edited nationally. This Network contains 1200 open access journals, which publish an average of 50 thousand papers per year.

By june 2019, SciELO Brazil was indexing 295 journals, which publish 21 thousand research papers per year. On the same month, its repository contained 350 thousand papers that, in 2018 alone, were accessed and/or downloaded in an average of 800 thousand times per months. Approximately 70% of SciELO’s journals are also indexed by Scopus or by Web of Science.

2.2. FAPESP Code of Good Scientific Practice

Another FAPESP milestone towards Open Science practices was the launching in 2014 of the Code of Good Scientific Practice, which, among other guidelines, stresses the need for open scientific procedures and open results of research funded by the Foundation. In particular, sections 3.2 and 3.3 of this Code are respectively dedicated to the "presentation of research findings and authorship", and the "recording, storage and accessibility of data and information". These sections specify that, after publication, research records must be made available to other researchers who may want to verify the study’s correctness or replicate or continue the study. Accessibility may only be limited for ethical or legal reasons.

2.3. Open Access to publications resulting from research funded by FAPESP

FAPESP’s funding of SciELO, as early as 1998, is an example of adoption of Open Access policies, before most such initiatives in the world that consider open access fundamental to the advancement of knowledge.

That was the first of a series of actions towards the same goal. Another example is the network of repositories of scientific papers, managed and maintained by the state’s public universities. This network was launched in 2008, ensuing a recommendation of FAPESP's Executive Board to that effect.

In the beginning of 2019, FAPESP formalized its Open Access Policy for Publications. Under this Policy, all journal papers that result from FAPESP-funded research must be made publicly available via their publication in institutional open access repositories, as long as this does not violate the copyright rules of the journal in which the papers appear. Researchers are free to decide the publication venue, as long as they comply with this Policy.

2.4. FAPESP and Open Data

At the end of 2017, FAPESP institutionalized the adoption of Data Management Plans (DMP) upon submission of project proposals. Such Plans must describe the data that will be used and generated throughout the research, how these data will be made public, and for how long they will be preserved. Initially compulsory for Thematic Projects, DMP are now compulsory for project submissions in all funding lines.

Following Open Data recommendations, scientists must deposit the data produced by projects funded by FAPESP in institutional repositories where the data will be accessible, maintained and preserved according to the standards of each scientific domain, under the ethical, legal and security constraints defined by the project and its associated domains. Such data must be made as open as possible, while ensuring that these constraints are preserved.

In parallel to the establishment of its DMP policy, the Foundation instituted in 2017 a workforce to design and implement a state network of open research data repositories. This workforce was composed of representatives of the state’s six public universities (University of Sao Paulo, University of Campinas, Sao Paulo State University, Federal University of Sao Carlos, Federal University of ABC, and Unifesp) and of the Aeronautics Institute of Technology. In 2018, the workforce was joined by CNPTIA (the Informatics Research Center of the Brazilian Agricultural Research Corporation, EMBRAPA). This network, available since december 2019 at https://metabuscador.uspdigital.usp.br/, hosts scientific data from research produced by the network members in all scientific domains.

FAPESP AND OPEN COVID-19 DATA

Thanks to the availability of this open research data infrastructure, which took 2 years to design and build, FAPESP was able to give a prompt response to the pressing need for reliable data for research on COVID-19. In June 2020, it launched COVID-19 Data Sharing/BR, an open repository containing health data on patients who had undergone COVID-related testing in Brazil since the pandemic outbreak in the country. 

This repository was created as a result of the collaboration of FAPESP with the University of Sao Paulo, and three prestigious Brazilian healthcare institutions - Instituto Fleury, Hospital Sírio-Libanês and Hospital Israelita Albert Einstein. 

In its first stage, the repository made available curated data on 177,000 patients, together with information on primary endpoints and patient transfers, corresponding to almost 5 million records that were pre-processed to meet Brazil’s data protection laws. It will subsequently contain medical images of COVID-19 patients, and associated reports. Periodically updated throughout the pandemic, this represents an exemplary open science effort launched within a public-private partnership.

2.5. FAPESP and Open Computational Processes

There is a growing amount of research conducted with the help of computational processes – e.g., using scientific workflows, electronic notebooks, or pieces of software. The publication of such processes via open repositories fosters reproducibility. Many funding initiatives at FAPESP require that all software developed within a funded project be licensed as free software.

For instance, the FAPESP - Microsoft Research MOU (started in 2005) and the FAPESP program on eScience and Data Science (started in 2014) are examples of funding lines that require that software developed within projects they support must be free. Yet as another example, FAPESP supports a research center on open science and innovation in Medicinal Chemistry (CQMED), which is associated with the Structural Genomics Consortium.

Many other research centers and projects supported by FAPESP also practice Open Science, through open data and processes – e.g., BIPMed - Brazilian Initiative on Precision Medicine, which is part of the Global Alliance for Genomics and Health or the Interscity research initiative on Smart Cities.