Help Us Integrate GitLab and the Open Science Framework

Andrew SallansFor years, the benefits of open source code development have been self-evident to the software development community: Transparency leads to collaboration, and collaboration leads to better and more secure code. The scientific community is just starting to understand these benefits.

The growing open science movement is using these same lessons to make the scientific process more transparent, so that research findings will be more reproducible. In order to realize the benefits of open science, we must use a wide set of research tools to enable transparency, which will lead to increased discoverability, reuse, and collaboration.

To that end, the Center for Open Science (COS) is funding the development of an integration between GitLab and the Open Science Framework (OSF), and is seeking interested members of the open source community to contribute to this effort.

A harsh reality

As with open source software, scientific collaboration leads to advances in knowledge. But there is a harsh reality: Scientists face tremendous pressure to publish results that are positive and innovative.

Because of this pressure to publish, research that does not result in significant findings often gets "lost" in the file drawer. This biases the research literature, making it look more beautiful than reality. The incentives for scientists tend to focus only on the publication.

But as David Donoho notes, the publication is advertising for the research, not the research itself. The research, including the methodology, collected data, analysis code, and workflow to generate that content, is closed, ignored, forgotten, or lost. This may be a significant contributor to the low rates of reproducibility of published scientific research.

This is changing, however. Funders, societies, and other organizations are recognizing the value of openness across the entire research lifecycle. Also, many commercial and non-profit groups are building new tools and services to make it easy and desirable to share research content and workflow.

A new infrastructure

COS, a non-profit science and technology company located in Charlottesville, VA, is building free, open source infrastructure as public goods to support the open science movement. With a mission to increase the openness, integrity, and reproducibility of research, COS supports communities that embrace open science practices, conducts metascience research, and develops and maintains free, open source software—principally the Open Science Framework (OSF), a web application that supports project management, data and materials archiving, and maintaining versioned histories of the entire research lifecycle. Simultaneously, the OSF integrates public and private workflows making it very easy for researchers to "go open" by sharing data and project materials with others.

The OSF features include collaborative wikis, automated document versioning, fine-grained privacy controls, and the ability to fork projects. The OSF meets researchers "where they are" by connecting to tools they already use via an abstracted API. As an application framework, the OSF brings the services researchers used together in a single place so that they can focus on their science and become more efficient with their workflow at the same time. Currently, the OSF connects storage providers Dropbox, Amazon S3, Google Drive, GitHub, Dataverse, and citation management tools Zotero and Mendeley.

Hundreds of other tools and services exist across the research life cycle, and researchers can use these to manage projects, collect and analyze data, keep track of manuscripts, and review and submit articles. COS is supporting many open source contributor teams to connect other services via the abstracted API such as: Vivo (search and discovery tool), DMPTool (data management plan tool), Evernote, Galaxy, PsychoPy, and Open Sesame—acquiring materials and collecting data, JASP (analytics tool), OneDrive and ownCloud (storage services), Zenodo, Dryad, and ShareLaTeX (repositories and authoring tools), and Open Journal System (OJS) and Ubiquity Press (publishing platforms).

Moving forward

Expanding connections to other tools helps scientists increase the effectiveness, efficiency, and reproducibility of their research. OSF users have repeatedly requested integration with GitLab, and we are seeking developers to create this connection.

We also encourage members of the open source community to explore the OSF and its repos for additional integration and development possibilities that serve the open science movement. If you or someone you know is passionate about open source development or shares our mission of improving science, we are actively seeking people to use these funds or to join our team of developers.

COS also offers a selective paid internship program where software developers gain valuable career experience working on free, open source software as part of a team committed to making a difference in the scientific community. Many interns work on user-facing issues of the OSF including integration with third party services. COS is committed to expanding on its mission of increasing openness, integrity, and reproducibility of research, with development of free tools and services for scientists. Broader involvement from the open source software community will broaden and improve the existing effort.