Jonathan Frazer

The 2026 Probabilistic Modeling in Genomics (ProbGen) meeting will be held at UC Berkeley, March 25-28, 2026. We have an amazing list of keynote speakers and session chairs:
probgen2026.github.io

Please help spread the news.

Borzoi-informed fine mapping improves causal variant prioritization in complex trait GWAS Genome-wide association studies (GWAS) have identified thousands of trait-associated loci. Prioritizing causal variants within these loci is critical for characterizing trait biology. Statistical fine...

I'm excited to share work on a research direction my team has been advancing: connecting machine learning derived genetic variant embeddings to downstream tasks in human genetics. This work was led by the amazing Divyanshi Srivastava! www.biorxiv.org/content/10.1...

*Easter egg alert* NOT in the published paper. We also benchmarked Evo 2 and while it did better than other gLMs (consistent that scale can improve gLMs), it still falls short of a basic CNN trained using one-hot sequences and far short of supervised SOTA

Excited to be presenting my paper "Deep Learning is Not So Mysterious or Different" tomorrow at ICML, 11 am - 1:30 pm, East Exhibition Hall A-B, E-500. I made a little video overview as part of the ICML process (viewable from Chrome): recorder-v3.slideslive.com#/share?share...

EurIPS is coming! 📣 Mark your calendar for Dec. 2-7, 2025 in Copenhagen 📅

EurIPS is a community-organized conference where you can present accepted NeurIPS 2025 papers, endorsed by @neuripsconf.bsky.social and @nordicair.bsky.social and is co-developed by @ellis.eu

eurips.cc

A Collectivist, Economic Perspective on AI Information technology is in the midst of a revolution in which omnipresent data collection and machine learning are impacting the human world as never before. The word "intelligence" is being used as...

What if AI isn’t about building solo geniuses, but designing social systems?
Michael Jordan advocates blending ML, economics, and uncertainty management to prioritize social welfare over mere prediction.
A must-read rethink.
arxiv.org/abs/2507.062...

Check out our latest work on the evolution of animal genome regulation out today in @nature.com. Nicely summarized below by @ianakim.bsky.social.
www.nature.com/articles/s41...

This is a major output from our ERC-StG project Evocellmap @erc.europa.eu at @crg.eu

One thing that really bothers me with the new "virtual cell" terminology is that it is currently largely focused on a very narrow definition of models that can predict effects of trans perturbations (gene dosage, drugs etc) on gene expression. 1/

New preprint - collab with the groups of @mehurles.bsky.social and @dr-appie.bsky.social. We imputed missing fluid intelligence test scores into ~170k @ukbiobank.bsky.social indivs & showed how this reduces bias and increases power for rare+common variant analyses. www.medrxiv.org/content/10.1...

🚨 New paper 🚨 RNA modeling just got its own Gym! 🏋️ Introducing RNAGym, large-scale benchmarks for RNA fitness and structure prediction.
🧵 1/9

There are many domain-specific noise processes for discrete diffusion, but masking dominates! Why? We show masking exploits a key property of discrete diffusion, which we use to unlock the potential of those structured processes and beat masking! w/ Nate Gruver and @andrewgwils.bsky.social 1/7

It's great to see some new notices of funding opportunities from NIH, but here's a reminder that Harvard researchers remain ineligible to apply for any federal grants.

"Harvard should no longer seek GRANTS from the federal government, since none will be provided."

www.harvard.edu/research-fun...

“The search is on for 25 researchers from around the world in Lund University’s largest ever international recruitment drive. The main focus will be on AI research and other areas in which the University has a strategic focus.”

www.lunduniversity.lu.se/article/lund...

🚨 Internship Alert! 🚨
Exciting opportunity @Genentech for a Computational Biology intern in Aging Research! 🧬 Work on #single-cell transcriptomics & imaging data to advance #aging biology insights with @mkoutrouli.bsky.social & me. Applications due June 26: lnkd.in/dBE28Wtk

✈️ 𝐓𝐫𝐚𝐯𝐞𝐥 𝐆𝐫𝐚𝐧𝐭𝐬: 𝐁𝐫𝐢𝐧𝐠𝐢𝐧𝐠 𝐆𝐥𝐨𝐛𝐚𝐥 𝐕𝐨𝐢𝐜𝐞𝐬 𝐭𝐨 #ESEB2025

🌍 At #ESEB2025, we believe great science should be accessible to all, regardless of geography or background.

📅 𝐄𝐱𝐩𝐥𝐨𝐫𝐞 𝐚𝐯𝐚𝐢𝐥𝐚𝐛𝐥𝐞 𝐠𝐫𝐚𝐧𝐭𝐬:
 🔗 https://lnkd.in/dXRzeUks

End-to-end differentiable homology search for protein fitness prediction.

@yaringal.bsky.social @deboramarks.bsky.social @pascalnotin.bsky.social

arxiv.org/abs/2506.089...

Average zero-shot mutation effect predictions across related sequences to reduce noise and improve accuracy.

@cwjpugh.bsky.social @paulinanunezv.bsky.social @jonnyfrazer.bsky.social Mafalda Dias

www.biorxiv.org/content/10.1...

Also, Charlie's #VariantEffect25 talk is available here www.youtube.com/watch?v=johJ...

Reminder that the MSCA postdoctoral program exists. If you have a PhD and want to work in a European lab, you have until September to apply. Just contact them now.

ec.europa.eu/info/funding...

📣 Latest from the lab: Performance of deep-learning-based approaches to improve polygenic scores www.nature.com/articles/s41...

Its thought deep learning will substantially improve PGS but the reality is MANY have tried but no/little gain has been seen so far. Here we report our negative results.

Population-scale inheritance analysis of 858,635 individuals reveals North Sea migration from the Middle Ages to the Industrial Revolution The North Sea's historical migrations have impacted the genetic structure of its neighbouring populations. We analysed haplotype sharing among 858,635 modern individuals from Denmark and Britain to in...

We have updated our preprint about 850,000 person UK-Danish haplotype sharing - sheding light on the rich history each person carries with them in their DNA and comparing this to historical records. Xiaolei Zhang postdoc in my group led this work. www.biorxiv.org/content/10.1...

Do protein language models store different structural elements in factorizable subnetworks?

To find out, we masked out PLM weights to suppress performance on CATH subcategories or secondary structure elements while maintaining performance on other sequences or residues.

The genetic architecture of an allosteric hormone receptor Many proteins function as switches, detecting chemicals and transducing their concentrations into cellular responses. Receptor switches are key to the integration of environmental signals, yet it is n...

1 plant hormone receptor ☘️
3,500 mutants, to single-site saturation 🧬
>45,000 binding and abundance measurements 📶

Very happy to present our latest work – where deep mutational scanning meets the world of small molecules.



www.biorxiv.org/content/10.1...



With @benlehner.bsky.social

[1/7]

I have an opportunity to hire a staff scientist for my lab. Looking for someone with outstanding skillset in ML/statistics, genomics applications; interest in mentoring, strong publication record, PD experience required.

Email CV to me+cc my assistant (see 'contact' on my website). Ad to follow.

We're thrilled to introduce PromoterAI — a tool for accurately identifying promoter variants that impact gene expression. 🧵 (1/)

If I was applying for PhD positions, Eli’s group would be at the top of my list.

NIH funding supporting the HMMER and Infernal software projects has been terminated. NIH states that our work, as well as all other federally funded research at Harvard, is of no benefit to the US.

Curiosity underlies a breakthrough in rare disease We must recognise and protect the pipelines that lead from research to real-world benefit

Both translational and fundamental curiosity-driven research are needed to fuel the incredible progress we're seeing in genomic medicine; an important message in this article and and some lovely quotes from @sarahlwynn.bsky.social
www.ft.com/content/25dd...

Thank you so much!

Want to improve your protein or genomic language model’s performance at zero-shot variant effect prediction? We propose a simple adjustment to likelihood-based predicton