Kaur Alasoo

First, a couple of weeks ago:

www.biorxiv.org/content/10.6...

Single cell RNAseq on blood from 200 donors across 4 sites in Indonesia spanning regional axes of genetic and environmental diversity. This is one of very few scRNA data sets from the Global South, including urban and rural settings.

Finally, today's offering! www.biorxiv.org/content/10.6...

This began life as a very different project which failed because we couldn't agree on defining eqtl sharing across cohorts. So two young members of the lab dug deeply into this - first @ijbeasley.bsky.social, then @patrickgibbs.bsky.social

I used Claude Opus 4.5/4.6 (and a bit of Codex GPT-5.3) to port edgeR to Python. See edgePython github.com/pachterlab/e...
This allowed me to develop a single-cell DE method that extends NEBULA with edgeR Empirical Bayes. All in one week. Details in doi.org/10.64898/202...

✨ A new star allele was born! Excited to share our first in vivo pharmacokinetic recall study involving 114 participants from @ESTbiobank with previously uncharacterised variants in CYP2C19 and CYP2D6:
www.nature.com/articles/s41... #Pharmacogenomics @pharmvar.org @clinpgx.org

✅ The UK Government has today granted access to half a million UK Biobank participants' coded GP data for health research.

This change will dramatically increase the power of UK Biobank's dataset to advance the diagnosis, treatment and management of conditions handled by GPs.

Hiring a PhD student! 🎓
Looking for a PhD student to work on the prediction of complex diseases by integrating genetics, multi-omics, and electronic health records. Details: genomics.ut.ee/sites/defaul...
Please share with anyone interested!

🧬 New preprint alert! After years of collaborative work across 52 datasets we are presenting eQTLGen phase 2: a genome-wide eQTL meta-analysis covering 43,301 blood samples: www.medrxiv.org/content/10.6... (1/8)

Thanks to amazing contributions of 52 cohorts we have been able to complete a full genome-wide trans-eQTL analysis in 43,301 blood samples. Many thanks to everyone involved, this has been a huge effort!

Preprint with the results of genome-wide eQTL analyses across 43,301 blood samples from the eQTLGen Consortium is now online! A major collaborative effort- many thanks to everyone involved, especially @robertwarmerdam.bsky.social, @ludefranke.bsky.social, and all eQTLGen collaborators!👇

I mean, death from infectious disease clearly has a genetic component related to the immune system… the distinction between intrinsic and extrinsic seems a little slippery.

Yeah, some of these 'extrinsic' causes of death are still partially heritable e.g. via risk-seeking behaviour. And the extent to which they are heritable could strongly vary between environments (e.g. how strong are social safety nets).

Every time someone uncritically repeats this meme about tens of thousands of undiscovered functional noncoding RNA genes in the human genome, an angel loses its wings.

I've started my own lab 🎉
PhD/postdoc positions available - reach out if curious about cerebellum evo-devo and autism spectrum disorders.

We’re based at Uni Tartu, Institute of Genomics (home to Estonian Biobank), and funded by @simonsfoundation.org @embo.org, and the Estonian Research Council.

Paper alert! 💻 How many cells do you need to train reliable deep learning models in regulatory genomics? We asked how data quality, sequencing depth, and dataset size affect training of sequence-to-function models from scATAC-seq. Out now www.nature.com/articles/s41...
(details below)

We have a nice example of convergence in this preprint as well:
pubmed.ncbi.nlm.nih.gov/40297438/

Branched-chain α-keto acid dehydrogenase complex + two regulators. BCKDK and BCKDK only dicovered with rare variants; PPM1K, DBT and DLD only discovered with common variants. All associated with BCAAs.

Also, gotta love Table 3. Apprarently they report an analysis across three cohorts, two of which don't even have the phenotype!

Sent!

Federated single-cell QTL meta-analysis reveals novel disease mechanisms #SingleCell 🧪🧬🖥️
https://www.biorxiv.org/content/10.64898/2026.01.20.700519v1

‘Shattered’: US scientists speak out about how Trump policies disrupted their careers Researchers lay bare the human toll of lay-offs, funding cuts and attacks on science one year after the president’s return to the White House.

Like so many, I’ve been affected by the new US federal stance against science. As a result of this and seeking a better life balance, I have accepted a new Full Prof position at the University Helsinki, Finland 🇫🇮🤩

I will miss CA deeply, but a new adventure awaits 🐟

www.nature.com/articles/d41...

Big congrats to @kmatlik.bsky.social and all the EMBO Installation Grantees 2026! 👏🎉

@burakyelmen.bsky.social from @estbiobank.bsky.social is hiring a PhD student!

Check out this work from Abdalla Alkhawaja et al in Karen Mohlke's group @klmohlke.bsky.social

Profiling RNA and ATAC of 68k+ nuclei from 39 individuals to better understand mechanism and context for GWAS loci

"Continued expansion of sample sizes, stratified analyses by ancestry, and careful consideration of lifestyle and other covariates will be essential to advance the reliability of peripheral epigenetic markers in psychiatric phenotypes like GAD."

Yes, that's a rare sight. With so many opportunities for confounding, I think it is genuinely surprising theat nothing came out as significant. I'm not sure I would be as optimistic about the future prospects as the authors are in the final sentece of the abstract, though.

But once again, the same trait can show very different sign bias across biobanks. Example: we estimate 72% of rare minor alleles (sign bias of 0.44) increase risk of type 2 diabetes in All of Us, while ~all increase risk in UK Biobank. Why? (11/21)

Works for me currently. Still down on your end?

Hi network 👋

⏰ Don’t miss it! #EMGM2026 in Davos (Switzerland) 🏔️
📅 14–15 April 2026
📝 Abstract deadline: 1 February 2026
✈️ Travel grants available
🎤 Amazing lineup of invited speakers
🔗 Details: emgm2026.sciencesconf.org

I went with a downhill skiing helmet with an integrated visor. Equally ridiculous.

Sled parking facility at our local Kindergarten in Estonia.

My own last, unreturned, correspondence with GISAID, after a months-long project building a version of Taxonium that would run within the walled garden (the dispute was over whether it could link out to e.g. the source code)