EVOLVEpro, our general AI model for protein engineering, is now out in Science! We show the ability to design and evolve dozens of unique proteins to have up to 500 times more activity or function using a combination of LLMs and regression models. 🧵👇 science.org/doi/10.1126/sc…

Introducing ESM Cambrian. Unsupervised learning can invert biology at scale to reveal the hidden structure of the natural world. We’ve scaled up compute and data to train a new generation of protein language models. ESM C defines a new state of the art for protein

Today we report that an engineered skin bacterium, swabbed gently on the head of a mouse, can unleash a potent antibody response against a pathogen. Could lead to topical vaccines that are applied in a cream. Djenet Bousbaine led the charge... nature 1/55

qPCR can measure expression of virtually any gene... in homogenized tissues. Here, we show an "in-vivo qPCR" equivalent, where we measure a sequence-specific transcript with synthetic serum markers and enzyme-based signal amplification. biorxiv.org/content/10.110…

Excited to finally share Perceptein, a PERCEPtron made of proTEINs: science.org/doi/10.1126/sc…

In new work led by Aditi Merchant with Samuel King, we prompt engineer Evo to perform function-guided protein design with high experimental success rates, including designs that go beyond natural sequences. We also release SynGenome, the first AI-generated genomics database. 🧵 1/N

Some of the recent LLLMs and efforts briefly reviewed, linked

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.

I’m excited to share our significantly-updated preprint on de novo antibody design, where we now demonstrate the structurally accurate design of scFvs (in addition to VHHs) with RFdiffusion! biorxiv.org/content/10.110…

Officially Dr. Ying! Successfully defended my PhD @harvard this week, marking 5 incredible years with Gladyshev Lab. Deeply grateful to Vadim and the entire team for their mentorship on this journey, as well as all of my collaborators and peers! Excited for the next chapter!

So excited to share our new paper out today in nature. We repurposed R2 retrotransposons to create STITCHR, a tool that can make diverse edits in mammalian genomes including scarless multi-kilobase scale insertions nature.com/articles/s4158…

Glad to finally see this out! Check out the fantastic post by ⁦Chris Fell⁩ ! nature.com/articles/s4158…