This is the eighth state of bims statement (sobs). Sobs comes out every 30 of January. It commemorates the first sober meeting between Gavin and Thomas on 30th January 2017. This statement contains a part by Gavin and a part by Thomas.

Gavin writes:

It has been another slow year of recruitment for selectors. We reached a high of 126 reports late in 2025. Despite reaching out to many potential selectors, the conversion rate is low. This is likely due to our marketing of Biomed News. Our efforts to explain the benefits of Biomed News are not likely to persuade someone from moving away from their current process. I think our long-term users are probably the best messengers for the platform. They use the platform and can explain the advantages with first hand experience.

Despite our slow recruitment, the number of subscribers is encouraging. Leading the way is Christian Frezza with 144 subscribers to his metabolism in cancer report. He is followed by Farhad Shokraneh with 122 subscribers with his AI in evidence synthesis report. This indicates that there are people who find these reports useful and could potentially help with our publicity.

Several of our selectors are active on social media and share their issues there. Along with Christian, Dario Brunetti and Catalina Vasilescu often reach close to 100 likes on their weekly posts. Thank you to these selectors for taking the time sharing these, it has helped to drive some interest and find potential selectors.

Over the summer of 2025 I attended the FEBS congress in Istanbul and presented a poster about Biomed News. This was an opportunity for present this to a large number of potential users. However, it was a disappointing level of interest, and nothing came from it. This further confirmed that presentation at scientific meetings is not an effective recruitment approach. We are open for suggestions to improve recruitment.

The highlight of the year was when Thomas visited Liverpool. It was an enjoyable and productive period where I got to learn what happens behind the screen of Biomed News and begin to learn some of the commands and organisation. We were able to discuss Biomed News at length and think of new ways for it to work.

We continue to maintain Biomed News as a free platform to use for biomedical research literature discovery to help the community that needs and wants to find relevant information efficiently.

Thomas writes:

The form to create reports has been up all this reporting time. It generates some request for opening reports, maybe like one times a month or so. But we seem to be having a hard time converting them into a least one report issue. My gut feeling is only one in three report requests yields at least one report issue. I am at a loss as to why. The most obvious would be that potential users don’t see the email. Another could be because they don’t read them. I am at a loss.

This year the Trump administration cast a shadow over the continued functioning of PubMed. I use the PubMed data files, as opposed to using the API. So far, the PubMed data files has generally been updated. The was a massive gap between 2025‒08‒20 and 2025‒09‒03. I got PubMed helpline to admit that the problem was at their side. This gap caused the first ever official issue delay. I had a lot of angst related to that.

Yes, it would be possible to use the API, but it is not clear whether this would be accessible from abroad. Our server is in Helsinki. The Trump administration did indeed stop the supply the data on related papers via the API for some time in the early part of the year. I only use this data at the opening of reports. I have a machine in the USA that I can deviate around for this purpose. But that machine is not big enough to run the entire processing chain of PubMed. For this occasional API call it should work. But re-engineering the PubMed processing to work with the API seems unreasonable give that the threat to the file data is low, and the cost of rewriting the entire data pipeline to go with the API is high. I suspect it could take me a couple of months to do that. And the risks are high to omit papers unless I keep the file-based system running.

On the topic of related papers, one other development has been to limit such papers at report opening. It turns out that if the seed papers has no MeSH heading attached then the related papers are pretty poor. I am grateful for advice for Marijane White of OHSU.

On 6 June NLM announced the availability of additional data that was not in the files but only in the API. Until that moment I was not even aware of a difference. At this time my theory is that the extra data is citation data. However, this seems usually only available for revised records. Thus when records come out, and we use them immediately, and the contain no citation data then we don’t have this anyway. So the policy will to stay to train from the records in the state that they came out in.

I made a trip to the UK. I met with Gavin 7‒20 to 7‒24. To raise to the occasion, I wrote checking software to for the report configurations. I am very grateful for having been hosted by Farhad Shokraneh 7‒24 to 7‒29. We worked an a service review of Bims for publication. Finally, I spend one night with Michael Upshall, who interviewed me last year.

At the end of the year, we moved server. This is for economic reasons. The Open Library Society needs the server for other purposes. So Gavin is again paying for a seperate server. The change over happened exactly at the start of 2026 in UTC. Moving server is a very complicated operation, compounded by the fact that we now used a cheaper network-attached storage. Some early users reported a problem with accessing the 2026‒01‒04 issue. I did find one mistake I made on the DNS.

Both write:

Once again, we thank you for your support, the time you spend to keep your reports up to date and your comments to improve the system. Wishing everyone a successful 2026.