End of year summary for 2025
We need to submit annual reports to Novo Nordisk Foundation as part of our grant agreement. Since I need to write something anyway, I might as well make it public and share it with anyone interested in our work over the years. It’s also a good way to reflect on what we’ve done and any challenges, some expectations for the upcoming year, and what we’ve learned. Plus, it’s easier to find when I have to make further reports over the years.
Major activities
This last year was an incredibly productive year for us in terms of tangible outputs. It felt like all our work building up the foundations in 2024 paid off in 2025 as we officially published two Python packages, three template tools, one R package, and developed a new workshop that we ran twice! Plus, we presented our work for the first time at a seminar and we hired another consultant to work with us on building software and creating teaching materials.
The two projects we’re collaborating with also started up in 2025, which lead to more collaborative activities. In particular, our initial work with them has been to help them set up their general digital infrastructure, their websites, and their data management plans.
Overall, 2025 was a very good year for us 🎉
Developed and ran a new workshop 📖
In early 2025, we received several requests to run a workshop on GitHub, since we use it almost exclusively in our work. So, we developed a workshop that introduces researchers to GitHub and how to use it for their projects. The workshop is a very gentle introduction to GitHub. In the end of May, we ran the workshop for the first time and in-person for staff, researchers, and doctors working at Steno Diabetes Center Aarhus. We had 24 participants attend the workshop. Then, we ran it again, this time virtually, for 11 PhD students and researchers who are or will be working on the DP-Next project (whom we collaborate with) in September. We received great feedback from both workshops, which I wrote a post about that also includes some figures of the feedback we received.
Presented at our first seminar! 🎤
Signe presented on our work in the Seedcase Project at a seminar, Open Science in Health Research: From Principles to Practice, in June. The seminar was hosted at Aarhus University and included researchers and practitioners on open science from across Europe. See her post about it, as well as the slides she used in her presentation.
Published Sprout! 📦
In July we officially published our first Python package, Sprout. Sprout is a package that helps create and manage a standardised and organised structure for storing and describing research data as a Data Package. I wrote a post about it that goes into more detail.
Published a template for Data Packages 🧰
Since we’ll be creating and developing many Data Packages for our collaborators, we decided to create a template tool to make it easier to create new Data Packages that follow the Seedcase structure. We also needed some way to keep these future Data Packages consistent with our current practices, and we do that by continually updating the template and synching it with the Data Packages created from it. So in the end of July we published our first template tool, Template Data Package. We published it so that others can use it to create a Data Package as well. I’ve written a post about it.
Published a template for websites 🧰
Across the Seedcase Project, we have nearly 11 different websites. Combined with the websites of our collaborators, which we also help build and maintain (e.g., the GitHub organisations onlimit-study, dp-next, and steno-aarhus, we have a lot of websites to manage and track. So, to make it easier to create new websites and to keep existing websites consistent with our current practices, we created Template Website, which we published in August. See my post for more details.
Published a template for Python packages 🧰
While we had only created and developed one Python package at this point, we knew we would be creating many more in the near future. As with the Data Packages and websites, we created and published Template Python Package in September to make this whole process easier for us. See my post for more details.
New team member! 👥
We hired a new consultant, Joel Ostblom, in October to work with us on developing our software and in creating teaching materials. He has extensive expertise from working as a professor at the University of British Columbia in Canada building and running data science courses, as well as being an active contributor to the open source packages altair and vega. He started working with us to contribute his expertise as we expand the number of software packages we build and as we shift to developing training materials for workshops. We’re excited to have him on board!
Published check-datapackage! 📦
We initially had planned four packages related to developing and managing Data Packages. However, in the process of building Sprout, we needed some way of checking that a Data Package complied with the Data Package standard. Since there were no similar packages available that did those checks, at least none that fit our needs, we decided to take a small detour to create and publish a Python package that fit our needs. We also knew that others might find a package like this useful, so this detour was as much for our needs as for others. So in the end of November we published check-datapackage, which ensures that a Data Package is compliant with the standard. See the post I wrote about this.
Published osdc R package! 📦
Since 2023, we’ve been collaborating with Anders Aasted Isaksen on converting his register-based diabetes classification algorithm into an R package so that others across Denmark can also use it. The package is used to classify diabetes status using the Danish registers. And finally, after a lot of work and many iterations, we officially published the Open Source Diabetes Classifier, osdc, R package in December 🎉 I wrote a post that has more details on this.
Collaborating projects officially started!
The two projects we are collaborating with, DP-Next and ON-LiMiT, officially started in 2025. While both had officially started in some way in late 2024, our actual collaboration with them began in September. And since then, we’ve been working with them to build up their websites, set up (some of) their digital infrastructures like with GitHub and GenomeDK, provide some formal and informal training, like the Introduction to GitHub workshop, and help with their data collection, processing, and management plans.
Challenges faced
We haven’t had any major challenges to deal with in 2025. The biggest “friction” points were mainly around coordinating and prioritizing work between our own internal projects and the work we do for our collaborators.
Managing more projects and being firm in prioritisation
As we officially started our collaborations with DP-Next and ON-LiMiT, there were a lot more projects and tasks to keep track of and manage. At least my time spent in planning and organizing our work increased quite a bit in 2025. We’ve had to be firmer with how we prioritise work and keeping our focus on the most important tasks at any given time.
This increase in projects and tasks has also caused us to go from working as a single team within the Seedcase Project to now being a group made up of several sub-teams working on different projects. This required us to change how we managed, communicated, and tracked tasks and work, which was a bit of a source of friction and trial-and-error until we found some processes and workflows that worked for us. We’re still refining these workflows and expect that we’ll keep developing and revising them the forthcoming year.
Summary
2025 was a great year for us in the Seedcase Project with many tangible outputs! And we’re excited about what 2026 will bring, especially as our collaborating projects start collecting data and we can start helping them more with data processing and organising, and in developing the remaining planned Seedcase Python packages. What’s the next steps for us? Check out our high-level roadmap as well as our more detailed project roadmap (which changes more often than the high-level one).