Open Source in Pharma: J&J's 5-Year Journey to R-Based Regulatory Submissions

Pharmaceutical companies face numerous challenges when transitioning to open-source solutions for clinical trials. Typical pain points are validation requirements, regulatory concerns, and organizational resistance. 

Johnson & Johnson is no exception.

Since 2019, Johnson & Johnson's Clinical and Statistical Programming team has been transitioning to open source. In the process, they've built an infrastructure and processes to leverage R and open-source tools within their highly regulated environment. Their journey reflects a thoughtful approach to innovation while maintaining the rigorous standards required for regulatory submissions in pharma.

This transformation wasn't achieved overnight. It evolved through careful planning, leadership support, and collaboration with industry consortiums and open-source communities. Starting with experimental applications and training initiatives, the team gradually expanded their capabilities to include validated package repositories, interactive frameworks, and ultimately, regulatory submissions created with R.

The J & J team offered valuable insights in their recent video on Posit PBC YouTube channel. If your organization is in a similar transition period - or thinking about it - reading or watching their story will demonstrate how building an open-source mindset within a large pharmaceutical company can foster innovation while ensuring compliance with industry standards.

Table of contents

• Introduction to J&J's Open Source Journey
• J&J's R Journey Roadmap (2018-2025)
• How J&J Built an Open-Source Culture
• Lessons Learned While Transitioning to R and Open-Source
• J&J's Best Practices from Years of Experience
• Future Opportunities
• Q&A Highlights
• Conclusion

Introduction to J&J's Open Source Journey

Johnson & Johnson's open-source journey began with a simple goal - modernize clinical trial operations while maintaining strict regulatory compliance. Suesh Kalal, who leads the Technology Solution Group in J&J's Clinical and Statistical Programming team, launched this initiative to bring innovative technology to their complex portfolio needs.

Leadership support proved crucial to the project's success. As Kalal explained during the talk, "When there is a great idea, it won't move forward without the support from leadership." The team built strong partnerships both internally (particularly with IT and statistical groups) and externally with industry experts and open-source communities.

When the team pitched their vision to senior management in 2019, they had a clear roadmap – but reality rarely follows perfect plans. Team members described their experience as a "bumpy ride" requiring quick adjustments and collaborative problem-solving. Despite these challenges, they progressed from initial implementation to regulatory submissions using R within five years, a key milestone in their 7-year journey.

What's refreshing about J&J's story is their transparency about both wins and struggles. It offers realistic insights for other organizations considering similar transitions.

‍Adopting open-source in pharma comes with challenges and opportunities. Read our guide for pharma leaders to find out what these are.

J&J's R Journey Roadmap (2018-2025)

Transition to open-source didn't happen overnight. In this section, we'll walk you through a 7-year-long journey from initiatives to achievements and future plans. 

2018: First J&J Shiny day

The seeds of J&J's open source journey were planted in 2018 with their first "Shiny Day" – a one-day conference to showcase R Shiny capabilities. Nick, one of the team members, recalls how this event sparked their interest: "I attended this and it was my first experience with R and Shiny. I left absolutely hooked with ideas buzzing in my head about how we could apply this in our stat programming group."

2019: Exploration and experimentation

The team moved from inspiration to action in 2019. They focused on what R could do for their specific needs and built their first Shiny applications for operational tasks. These initial projects didn't handle clinical study data but instead transformed Excel-based processes into more efficient web applications. Success came quickly, but it brought a challenge of not having enough skilled people on board to meet growing demands.

2021: Becoming a flagship initiative

By 2021, the R initiative gained enough momentum to become a flagship project within the organization. This status brought more resources, investment in infrastructure, and IT support. The team credits their IT colleagues, especially Satish Mury, for their critical role in building the compute environment needed for this work. They also expanded their external connections, joining industry consortiums and learning from other companies' experiences.

2022: First open-source package

The team hit a major milestone in 2022 by releasing their first open-source package. Nick describes this achievement with pride: "Looking back at 2018, being a package developer and open-sourcing something seemed impossible! We were super excited." They also created code catalogs to help their statistical programmers apply R to everyday tasks and selected early adopters to test these tools before wider release.

2023: Focus on submissions

In 2023, the team tackled what Nick called "the elephant in the room" – regulatory submissions. While confident based on industry initiatives and external efforts, they hadn't yet used R for their own submissions. The team focused on creating internal processes for R-based submissions and invested more in their GxP compute environment for interactive applications.

Learn more about the importance of GxP from Appsilon resources:

• Regulatory Compliance for R- and Python-based Scientific Computing Environments
• FDA Compliance in Software Development: Cases Where Poor Software Quality Led to Costly FDA Rejections
• How a GxP Audit Will Help You Go Through FDA and EMA Submission

2024: Hybrid submissions achieved

The team's efforts paid off in 2024 with their first four hybrid submissions to regulatory agencies. These submissions used SAS for datasets and R for outputs (tables, listings, and graphs). Nick explains: "The elephant was no longer in the room. We completed our first hybrid system submissions in 2024." The team also shifted their strategy from developing J&J-specific tools to adopting and adapting existing open-source frameworks.

2025: Future plans

Now in 2025, the team works to make R "business as usual" for their static outputs. They've expanded their focus to include admiral, an open-source package for clinical data creation, and adopted interactive frameworks with their first users already testing J&J versions of these tools.

How J&J Built an Open-Source Culture

The transition to open source at J&J wasn't just about technology - it required a new collaborative mindset. Their approach balanced community participation with the strict requirements of pharmaceutical research.

Benefits of open-source collaboration

J&J discovered multiple advantages as they embraced an open source mindset. Tad, one of the team leads, explained how they benefited from "thousands of hours already spent on frameworks or other solutions" they could reuse and build upon. This approach freed their team from reinventing tools that already existed in the community.

They also found value in making the existing solutions more robust through their contributions. Most importantly, open source enabled knowledge exchange across the industry. Tad posed an interesting question during the presentation: "Are we as pharma companies doing the same efforts but differently?" This highlighted the opportunity for standardization across the pharmaceutical sector.

Examples of successful implementations

The J&J team shared two specific examples of their successful open source collaborations. The first involved implementing True Font and RTF for R Tables. Their internal team adapted these technologies to match the exact formatting requirements needed for their clinical trial reports. In short, it allowed them to produce outputs that maintained consistent appearance across different document formats.

The second example demonstrated how they extended interactive capabilities with a Transform Module Output. This enhancement, released just two weeks before their presentation, let teams to add custom elements to outputs generated by existing modules without rebuilding them from scratch. The feature allows users add footnotes and decorative elements to adapt standard visualizations and reports to their specific needs.

Customization and extensibility

A major benefit of the open source approach was the ability to customize solutions for J&J's specific requirements. Tad demonstrated how the same base module could display different outputs based on custom settings – showing how one solution could adapt to various needs across departments.

"With your contribution, you might bring different new capabilities and functionalities that we all can build into one solution that fits all of us in the future," Tad explained. This vision extends beyond simple modifications to imagining a future where open source tools evolve to serve the entire pharmaceutical industry with consistent but flexible reporting capabilities.

The culture shift at J&J moved them from working in isolation to active participation in a broader community. Their experience shows how large pharmaceutical companies can both benefit from and contribute to open source while still meeting strict regulatory requirements.

Lessons Learned While Transitioning to R and Open-Source

J&J's journey with R revealed key insights that can help other organizations navigate similar transitions. Each challenge they faced created opportunities to refine their approach.

In this section, we'll walk you through the key lessons learned in years long transition process.

Training and knowledge development

J&J invested in comprehensive training programs tailored to different audiences, from statisticians to programmers. They created diverse learning resources including online courses, workshops, and coding sessions to accommodate different learning styles. Their statistics colleagues played a vital role in customizing these training materials.

Infrastructure and scalability

A robust infrastructure formed the backbone of their transition to open-source. As R packages evolved and interactive use cases grew, J&J made sure their systems could keep pace with demand. Mark highlighted the IT team's contribution: "A big shout out to IT, especially Satish M's team, for their fantastic support in helping us make the right infrastructure choices."

Standardization

Consistent practices proved essential for smooth collaboration across teams. J&J implemented standardized frameworks, packages, and code repositories to improve both quality and efficiency. This approach meant teams worked with the same building blocks and tools, making knowledge transfer simpler and reducing redundant efforts.

Power of the open-source community

The R community offered J&J access to a wealth of packages, resources, and support. This ecosystem accelerated their learning curve and sparked innovation within their teams. Rather than building everything from scratch, they leveraged existing tools and contributed improvements back to the community.

Change management

Early adopters played a key role in driving broader acceptance of R and open-source. When they showcased the interactive potential of tools like Shiny, it sparked interest across the organization. 

J&J discovered that phased implementation worked best in their case. In other words, this means gradual rollouts for different compounds or studies, as it allowed for smoother transitions and iterative improvements.

Risk mitigation and scaling strategies

Pilot programs helped J&J identify challenges early in their journey. These test runs provided crucial lessons before full-scale implementation. The team emphasized the importance of integrating R into daily routines with continued support and resources to make it "business as usual" for their staff.

J&J's Best Practices from Years of Experience

J&J team distilled their experience into practical recommendations for organizations thinking about similar open source initiatives. These strategies helped them overcome common obstacles, and we'll share a couple of them in this section.

Training and knowledge development

Creating structured training programs tailored to different audiences proved essential. 

J&J designed their curriculum to evolve as new tools and techniques emerged within the R ecosystem. They also focused on training teams on standardized ways of working with R and specific packages, ensuring consistent practices across projects.

Culture and engagement

Fostering curiosity and an open-source mindset accelerated R adoption in the organization. The team encouraged members to explore both static and interactive R applications through hands-on learning. 

This approach helped them leverage community resources while also contributing back to that community. Mark emphasized that "embracing an open source mindset allows us to not only use available resources but also enhance collaboration and innovation."

Documentation and infrastructure

J&J documented which validated packages existed and created knowledge-sharing platforms. Their "R Portal" served as a one-stop shop for all information related to their R journey, and has made resources easily accessible to everyone involved.

Feedback and continuous improvements

Creating feedback loops helped refine processes based on real user experiences. The team implemented mechanisms to collect input and monitor R adoption rates across different groups. These metrics guided their approach to training and support, allowing them to address pain points quickly.

Demonstrations and adoption

Compelling use cases highlighted the benefits of R for specific business needs. J&J showcased both static and interactive projects to inspire broader adoption, and has demonstrated how these approaches complement each other. 

Their "Shiny Days" events gave teams opportunities to share successes and inspire others through practical examples.

Change management

Planning for organizational change proved as important as technical implementation. 

The team developed strategies to address resistance and proactively respond to concerns. Mark noted that "success comes from ongoing efforts and resilience. Every step gets you closer to your goals, so embrace the challenges as key building blocks on your road to success."

Future Opportunities

J&J team is proud of their progress, but they remain focused on new opportunities that will further enhance their R implementation. They identified several promising directions moving forward.

Expanding external collaboration

The team plans to deepen their involvement in industry-wide initiatives. Nick highlighted their participation in a PHUSE project focused on "teal enhancements for cross-industry adoption." 

This collaboration will showcase interactive capabilities to regulatory agencies through presentations and workshops at the first EU version of CSS (Computational Science Symposium). These efforts aim to demonstrate how interactive tools can support regulatory review across the pharmaceutical industry.

Supplemental interactive submissions

J&J sees potential in using interactive tools to enhance regulatory submissions. Nick emphasized that their initial approach focuses on supplemental rather than replacement content: "When we think about submissions, it's not necessarily replacing things but adding tools to help reviewers quickly understand the story." Their plans include standalone HTML files and Shiny applications developed through the R Consortium's Pilot 4 working group initiatives.

Generative AI integration

The team has begun exploring how generative AI can support statistical programmers within their existing workflow. 

Rather than pursuing broad AI applications, they focus on building internal modules that follow J&J standards to provide quality responses to their statistical programmers. Nick described this approach as creating a "personal SME that is never too busy to answer your question" with immediate, reliable guidance. The team will present their progress at upcoming industry events, including a panel discussion on "GenAI for code generation" at the PHUSE CSS event.

Q&A Highlights

The presentation concluded with questions that revealed additional insights about J&J's experience. The responses you'll see below addressed common concerns organizations face when implementing open source tools in regulated environments such as pharma.

Package validation

Package validation showed to be a top audience concern. The J&J team partnered with a vendor to develop a streamlined validation process. "We built a process with the vendor and our IT team to create a CI/CD pipeline for validation," Suesh explained. Their approach involves releasing two containers per year with validated packages, which balances innovation with compliance requirements.

FDA submission experience

When asked about regulatory acceptance, Nick shared that while their submissions were still under review, they had received "no findings that have come back from regulatory questioning our use of R." 

The team submitted four hybrid submissions in 2024, using SAS for datasets and R for outputs. While they couldn't share specific therapeutic areas due to confidentiality, they promised to share outcomes once the submissions completed their review cycle.

Infrastructure details

J&J built their R environment on AWS Amazon Workspace with containerized Posit Workbench and Connect deployments. This containerized approach allowed them to manage validated packages through controlled releases while still maintaining a development area where users could explore packages from CRAN. 

For those seeking more details, they referenced previous webinars by their IT lead Satish Mury that provide deeper technical information.

Data integrity and reproducibility

Maintaining data integrity and reproducibility relies on their containerization strategy. "Packages come in, we evaluate the risk, mitigate it, and put testing and paperwork in place," as Nick explained. 

These containers preserve the exact environment needed to reproduce outputs at any point in time. Mark added: "For reproducibility, you need environment, data, and code. Within our environment, all these elements are validated and preserved."

ROI and business outcomes

While specific ROI numbers weren't shared, the team emphasized substantial benefits from their open source approach. 

Suesh noted: "The return on investment with open source is huge, with opportunities we're still exploring." He highlighted value derived from community support, collaboration through consortiums like R Consortium, PHUSE, R in Pharma, and Pharmaverse, all working toward common goals of promoting R in regulated environments.

Training strategies

Effective training combined structured learning with practical application. The team observed that newcomers to the organization adapted quickly to R, especially those without prior experience with other statistical software. 

Mark recommended: "Let them experience R through interactive applications, provide adequate support, and work through use cases together." They emphasized community support, both internal and external, as critical to successful adoption.

He and the team emphasized that learning is only the first stepreal adoption happens when scientists gain hands-on experience using new tools in their daily work. Community support, both internal and external, was also critical to making the transition stick.

Conclusion

J&J's journey from initial experimentation to successful regulatory submissions demonstrates that open source adoption in highly regulated environments is both possible and packed with benefits. Their experience shows that with proper planning, infrastructure, and cultural changes, companies can leverage R and other open-source tools while maintaining compliance with strict regulatory requirements.

For pharmaceutical companies and other regulated industries, the move to open source represents more than just a technology shift - it offers a path to improved collaboration, innovation, and efficiency. The open source approach allows organizations to benefit from community-developed solutions while contributing their expertise back to the ecosystem.

As regulations evolve and data science continues to transform clinical research, companies that embrace open source position themselves to adapt more quickly. J&J's parting advice summarizes their philosophy: "If you want to go fast, go alone. If you want to go far, go together." Their journey proves that even in the most strictly regulated environments, open source can thrive when implemented with care, collaboration, and commitment.

If you have an hour to spare, we encourage you to watch J&J's full presentation on Posit's YouTube channel.

If you're a clinical manager leading a data department, check out our Definition of Done (DoD) checklist to ensure your team meets compliance and quality standards for FDA and EMA submissions.