Interesting that large-scale pretraining on eukaryotic DNA sequences yields so little, while a convolution-based model like Puffin can learn a lot from a single modality. Any middle ground🧐?

Join us in June for the 2025 SEED meeting! The organizing committee has some great speakers lined up! Abstracts are due March 23rd! synbioconference.org/2025

Engineered cells are finding their way into agriculture, living materials, consumer products and medicines. How do we design these cells and adapt regulatory rules to facilitate their safe and efficient use? #synbio #biotech rdcu.be/d8xLG

Looking forward to it!

Can DNA sequence models predict mutations affecting human traits? We introduce TraitGym, a curated benchmark of causal regulatory variants for 113 Mendelian & 83 complex traits, and evaluate functional genomics and DNA language models. Joint work w/ gokcen and Yun S. Song 🧵👇

We trained a genomic language model on all observed evolution, which we are calling Evo 2. The model achieves an unprecedented breadth in capabilities, enabling prediction and design tasks from molecular to genome scale and across all three domains of life.

Dive into our lab’s first peer-reviewed publication—co-led with our friends Ben Adler and Muntathar Al-Shimary from the Doudna lab! If you’ve already read the preprint, don’t miss the new updates: doi.org/10.1038/s41564… Nature Microbiology

We're partnering with LOTTE BIOLOGICS, a global CDMO, to scale therapeutics manufacturing. Our customers can now transition seamlessly from cell line development using our CHO EDGE system all the way to large-scale GMP production, with faster cycles. asimov.com/news/lotte-bio…

Thrilled that our work on this problem with Karel Břinda, Zamin Iqbal, and others is out in Nature Methods today! We used phylogenetic compression (described in the thread) to compress every microbe ever sequenced onto a flash drive so that it can be searched with a laptop!

Seeing what bacteria see from a flying UAV 👁️ rdcu.be/ehkQB #synbio #biotech Itai Levin Yonatan Chemla

Today, we are launching the first publicly available AI Scientist, via the FutureHouse Platform. Our AI Scientist agents can perform a wide variety of scientific tasks better than humans. By chaining them together, we've already started to discover new biology really fast. With

🚨 new paper alert! science.org/doi/10.1126/sc… During my PhD, one of the most frustrating challenges was trying to interpret genes labeled as “hypothetical proteins.” 1/n

1/6 Excited to share our latest preprint: "MORPH Predicts the Single-Cell Outcome of Genetic Perturbations Across Conditions and Data Modalities". 🔗 biorxiv.org/content/10.110… 🧵 👇 Here is what MORPH is in a nutshell!

In 1965, Margaret Dayhoff published the Atlas of Protein Sequence and Structure, which collated the 65 proteins whose amino acid sequences were then known. Inspired by that Atlas, today we are releasing the Dayhoff Atlas of protein sequence data and protein language models.

I’ve always been interested in the stories behind BioNumbers. Textbooks say DNA polymerase, for example, replicates 220 nucleotides per second. But how do we know this? How can one actually measure the speed of a single enzyme? A fast typist, for comparison, types ~80 words per

This is so interesting. The conversion of CS into biology continues.

Really glad to see the progress in this field! perhaps one or two references could be added? 😉

The big one is finally out!! In this paper, we set out to provide insight into the fundamental question; How do the individual cells from complex tissues regulate their proteomes? Brief summary of our findings 👇 biorxiv.org/content/10.110…

We are excited to share GPN-Star, a cost-effective, biologically grounded genomic language modeling framework that achieves state-of-the-art performance across a wide range of variant effect prediction tasks relevant to human genetics. biorxiv.org/content/10.110… (1/n)

Excited to announce mBER, our fully open AI tool for de novo design of epitope-specific antibodies. To validate, we ran the largest de novo antibody experiment to date: >1M designs tested against 145 targets, measuring >100M interactions. We found specific binders for nearly half