Open Source in Regulatory Review: nlmixr2 Developers Meet Regulators
By Matthew Fidler in nlmixr2 rxode2 babelmixr2
April 17, 2026
The nlmixr2 development team, represented by Bill Denney, Lisa Hendricks, and I (Matt Fidler), recently gave a invited seminar to the FDA quantitative clinical pharmacology reviewers to discuss nlmixr2 and how open, community-built modeling tools can be helpful in regulatory settings. The conversation was constructive and forward-looking, and a clear theme emerged: anything that lets regulatory reviewers trust, reproduce, and exchange models across tools would be very useful.
As we go through these reflections, I must add that these are my takeaways and are not necessarily the views of the FDA, nor are they necessarily the views of the companies or individuals represented by the nlmixr2 working group, or the R Consortium.
Now that the disclaimers are out of the way, here are some highlights from the discussion.
Participants emphasized that FDA is tool-agnostic where submissions are concerned and have accepted several based on work done in nlmixr2 (as well as many other tools). This standpoint, however, necessitates a lot of cross-conversion between tools, since not all common tools are available for regulators to use.
The following specific points came up:
Participants were excited about model interchange and what it could unlock for reviewers and sponsors.
Interest was expressed in the eventual ability to interchange between
any model format (including NONMEM (nonmem2rx), Monolix (monolix2rx),
Campsis, dMod, mrgsolve, Pumas, Phoenix, PKsim, SimBiology, Ubiquity
and many many others). While the nlmixr2 dev team has no immediate
plans to support additional formats currently, we understand the gap,
and would love to have this kind of functionality available. nlmixr2
is open source, so if anyone wants to add this, we invite you to have
at it. We look forward to your contributions!
Interoperability, transparency, and reviewer-friendly workflows are
crucially important. One of our primary focuses moving forward is to
strengthen nlmixr2s regulatory readiness through clear validation
signals, smooth model format interchange, and providing better
training resources.
So, then:
What does “regulatory readiness” mean for nlmixr2?
It means our tools and workflows support reproducible results, clear documentation, and reviewer-friendly outputs, so regulators can confidently follow what was done and why, without ambiguity.
Has nlmixr2 been used in regulatory submissions?
Yes, and no issues were raised about its use.
Does the FDA require any specific modeling tools?
No. A key point, emphasized repeatedly, was that the FDA is tool-agnostic. The emphasis is on scientific validity, clarity, and reproducibility - so fitness for purpose, not branding. While the common wisdom in industry is that certain tools are approved by the FDA, being tool agnostic has in fact long been the agency’s position.
What is “model exchange” (or model interchange), and why is it important?
Model interchange is the ability to move a model between tools without losing meaning or correctness. This is important, because:
- Make reviews more consistent
- Reduce translation errors
- Support reviewers’ ability to work in the tools they know best
Why does “full model export” matter?
Reviewers need to see the complete, self-contained model definition. Relying on separate libraries or implicit dependencies can make it harder to validate and/or compare what was actually run, especially if reviewers are not familiar with specific tools.
Summing up
Overall the nlmixr2 team enjoyed the conversation and appreciated the
feedback from the colleagues at the FDA, and we are looking forward to
continued interaction with the FDA.
By the way, the slides shared with the FDA are found in our courses.
- Posted on:
- April 17, 2026
- Length:
- 3 minute read, 562 words
- Categories:
- nlmixr2 rxode2 babelmixr2
- See Also: