Shai Carmi

@shaicarmi.bsky.social

Associate professor at the Hebrew University of Jerusalem. Statistical, population, and medical genetics; preimplantation genetic testing. Views my own. http://scarmilab.org

Come and join our dream team! We’re recruiting a PhD candidate with a strong background in statistics, machine learning and (human) genetics to work at the interface of biological sequence foundational models and scores polygenic score methods.

Scaling genetic clustering (ADMIXTURE/STRUCTURE-style unsupervised clustering) is now easy. Out in Bioinformatics, our ADAMIXTURE method clusters all of UK Biobank (500k genomes) in just over a hour with likelihoods as good as ADMIXTURE, which would take weeks. doi.org/10.1093/bioi...

ADAMIXTURE: adaptive first-order optimization for biobank-scale genetic clustering

AbstractMotivation. Estimating genetic clusters from sequencing data is a fundamental task in population and medical genetics, enabling demographic inferen

doi.org

A meta-science thought. There are >11k papers on PubMed on polygenic scores. It's a huge waste of resources. It's mostly one question that matters. When reporting to people either their real PRS or a random PRS, do people in the real PRS group have better health outcomes years later?

Bild

Pre-print on the impacts of the second plague pandemic and concurrent historical processes in 529 human genomes (mean 8.8x) from @ucph.bsky.social, Amgen/deCODE genetics, and many collaborators. We'd be grateful for any feedback before publication, so please share!

PCA of ancient individuals included in our study. We observe mostly northern European ancestry with three notable outliers.Selection scan in pre-Black Death versus post-Black Death samples showing no significant results.
bioRxiv Genomics@biorxiv-genomic.bsky.social · last mo.

Genomic impact of the second plague pandemic on three human populations https://www.biorxiv.org/content/10.64898/2026.06.29.730585v1

The UK Biobank being inaccessible for half a year is annoying but also a great natural experiment. I'm wondering if in 3-4 years scientists could see the impact. Could teach us a lot about the scientific process and the value of biobanks.

All talks are now available including mine: www.youtube.com/watch?v=zZmF... Luckily the video became online on the same day as the paper I talked about at the end of my talk "confidentially" 😅

Du Paléolithique moyen au Paléolithique supérieur... (4) - Jean-Jacques Hublin (2025-2026)

YouTube video by Sciences de la vie - Collège de France

youtube.com

Collège de France@college-de-france.fr · last mo.

#Science #Paléoanthropologie 💀 « Sapiens remplace Néandertal » 🖥️ Les vidéos du colloque organisé par le  Pr Jean-Jacques Hublin, titulaire de la chaire Paléoanthropologie, sont disponibles. 👉 https://tinyurl.com/4pkf3nty

Accepted a review request two weeks ago for a Springer journal. Reminder emails on days 6, 9, 13, and again on day 13. Spent a day reading and preparing the review. On day 14, just before sending out the review, received a message that my review is no longer needed. Why do we agree to this???

Nice paper studying the extent to which an individual with known traits can be linked to their genome. Authors argue that the practical risk is minimal, except in specific settings where the number of candidate genomes is very small.

bioRxiv Genetics@biorxiv-genetic.bsky.social · 2mo ago

Evaluating anonymized genome re-identification using polygenic predictions and its implications for data privacy https://www.biorxiv.org/content/10.64898/2026.06.10.731306v1

Great to see @annikafreudiger.bsky.social paper published! 🎉 We analyzed how genetic markers are used to estimate relatedness in wild animals - a big field with 2861 relevant publications. 📈🧬 Even in 2026 mainly STRs are used - and we argue for low coverage WGS data and a simple aDNA estimator.

Molecular Ecology@molecology.bsky.social · 2mo ago

New in MEC! @annikafreudiger.bsky.social et al. assess genetic relatedness on rhesus macaques and find that low-coverage whole-genome sequencing offers an accurate, cost-effective alternative to widely used methods. @wileyecology.bsky.social 📷 Caribbean Primate Research Center 🔗 buff.ly/pPOONsu 1/3

Very excited about this preprint that we just posted. It introduces the Genomic-Relatedness Matched Association (GRMA) study. It’s an extension to family-based GWAS that uses extended relatives beyond only siblings in diverse-ancestry data with very little bias. (Link below.)

Bild

New preprint! Introducing DIMPLE-GWAS - a new dimensionality-reduction method to capture latent endophenotypes. Applied to pleiotropic brain imaging data in UK Biobank, we developed a new brain atlas composed of 25 core "clusters" Details below and at my #SOBP talk Sat 3pm

Bild