Maren Buettner

Absolutely nothing to add. Somehow became my fear whenever I submit a method paper. I do understand the legitimacy und utility of limited benchmarks in the paper and totally agree that extensive benchmarks should be a community effort. Usability besides performance is also key.

Hot take. Excessive benchmarking demands on computational method papers slow down progress. Benchmarks in the paper need to show that your solution works well enough for your goals - not that it is the best solution ever proposed or how exactly it differs from all other existing solutions.

BIG ANNOUNCEMENT📣: I haven’t been this excited to be part of something new in 15 years… Thrilled to reveal the passion project I’ve been working on for the past year and a half!🙀🥳 (thread 👇)

🌟 Thank You to Our Silver Sponsor: Calico Life Sciences ! 🌟

We're excited to recognize Calico as a Silver Sponsor of scverse conference 2025! 🥈

Details in thread! 🧵
#scverse2025 #AgingResearch #ComputationalBiology #SingleCell #Longevity #CalicoLabs #SpatialOmics

Join EMBL AI!

We have ambitious plans to transform life sciences research through AI.

Interested in using ML/AI to solve complex interdisciplinary challenges?

We’re looking for two senior leaders based at EMBL Heidelberg👇

Win for #openscience!

Consider attending the MOPITAS autumn school on Spatial Transcriptomics in Munich with expert lectures and hands-on sessions for all steps of spatial transcriptomics data analysis.
event.sdu.dk/mopitas-scho...

scverse conference 2025 Follow us on our channels to learn more details in the coming weeks

📣 Mark your calendars! The 2025 edition of the scverse conference will take place on 17-19 November at Stanford University (US) scverse.org/conference20...

Call for abstracts and registrations coming soon!

Thank you for following my journey on DNA sequencing. I won't share any other personal health related insights from the DNA sequencing results here. Let me know what you think and if you have any other questions, comments, concerns. I'm happy to discuss.

There are clear limitations to the interpretation of SNPs on the individual level. There is already only limited insights on the population level in large cohort studies like UKBB and aging phenotypes. And don't get me started on SNP to gene to trait links, or significance, for instance.

So overall, I gained some peace of mind concerning my breast cancer risk and that was important to me. Some other things could not be explained from DNA sequencing alone like the lactose intolerance.

To close, I keep getting emails about my "optimal diet" for heart health etc. - I honestly think that such claims go beyond what we can currently infer from DNA sequencing data.

That also concerns some findings were I am supposedly a carrier of, like Hirschsprung disease - but then again, I am a healthy individual and these results may only become relevant once I or my family members develop a particular health issue.

The sequencing provider also provided an "AI report" to interpret my results. Overall, I don't think that I learned a lot from those, especially given my overall health status is great. It would be a different story I had a particular health issue. Then genetic testing could give add. insight.

Lastly, I sorted the results by risk frequency. The top hit is related to prostate cancer, so I think that I am in the clear.

Next, I wanted to see if any of my SNPs point towards the lactose intolerance. The report identified 61 SNPs related to this condition, but all were labeled as having no impact or harmless variant. So no explanation on this one.

In short, I am a middle-aged healthy female with lactose intolerance. My paternal grandmother died of breast cancer at a young age, so first I wanted to know if I carry a risk for hereditary breast cancer (pathogenic BRCA mutations) - I don't so that is great news.

Update on my DNA sequencing journey. I actually got results back after 8 weeks (mid March). It took me a while to process the results and think about the level of detail to share. So let's go.

Wegen Trump: Zentralbibliothek Medizin baut Alternative zu US-Datenbank auf Ein deutsches Informationszentrum will eine Alternative zur PubMed-Datenbank in den USA entwickeln, um die Informationsversorgung in der Biomedizin zu sichern.

Germany's central medical library, ZBMed, is building an alternative to PubMed. It will be open source, based on MeSH & associated data standards, built to be sustainable in case PubMed becomes unreliable - with APIs & MeSH maintenance. Open meeting called! 1/2 www.heise.de/news/Wegen-T... #MedLibs

🧠 Neuroinflammation: Research by DZNE, @unibonn.bsky.social and others suggest that an over-active immune system of the brain is a driving force in “spastic paraplegia type 15”, a rare neurodegenerative disease associated with movement disorders.

👉 Paper @jem.org: rupress.org/jem/article/...

Thanks for sharing! I share the sentiment that the single-cell foundation models still overpromise and underdeliver. IMO future benchmark scenarios should be more complicated and focus on perturbation predictions. I’d love to see a benchmark on the Tahoe-100M dataset next.

Feature selection methods affect the performance of scRNA-seq data integration and querying www.nature.com/articles/s41... 🧬🖥️🧪 github.com/theislab/atl...

March 21, 2-3 pm ET: What to do when your federal grant has been stopped or terminated

If you have a federal grant, I *highly* recommend going to this Friday's webinar on responding to grant terminations.

Lisa Brown, former general counsel at Dept of Ed, will present along with other lawyers.

Friday, March 21, 2-3 pm. More info & register here: www.linkedin.com/pulse/regist...

If you're in the US, considering Europe for your PhD, and you identify as a minority, I will chat with you and review any application documents you need 🧪

Friends: Science is under threat! Worldwide, forces seek to make it corporate-like, focused on deliverables, or even the enemy of society. It's time to take a stand to celebrate the science we know and love.
Introducing: Woodstock.Bio2 + Night Science
June 10-13, Prague & Woods of Bohemia
Join us!

Release notes Version 1.11: 1.11.0 2025-02-14: Release candidates: rc2 2025-01-24, rc1 2024-12-20. Features: rc1 sample() supports both upsampling and downsampling of observations and variables. subsample() is n...

🎉 Scanpy 1.11.0 is out! 🎉 just after reaching 2000 stars on GitHub!

- sc.pp.sample replaces subsample with many new features
- Sparse Dask support pca
- session-info2 package for more reproducible notebooks

See the release notes:

So my samples passed QC and made it into the sequencing. The vendor's web app cautions that this process may take up to 8 weeks. Oh well. Hang in there.

That’s an interesting perspective. All model systems come with pros and cons. The translation to an in vivo (human) system is a challenge, agreed, but we can manipulate cell lines/organoids more efficiently and systematically than in vivo systems, resulting in less confounded data.

One week after sample collection, my samples have arrived in the lab and are now being processed. Fingers crossed that it passes the QC.

The tips of the sample swabs then came to rest in the tubes, and I packed them in the provided return envelope. I slipped the envelope in the mailbox to have it shipped to Texas, and I received a couple of status updates via email.