Foundation models for single-cell biology should be zero shot, drawing value from emergent organization. We present Universal Cell Embeddings🌌, which can represent cells from any tissue/species in a fixed space Yusuf Roohani @aggwall Jure Leskovec Stephen Quake biorxiv.org/content/10.110…

The Gemini era is here. Thrilled to launch Gemini 1.0, our most capable & general AI model. Built to be natively multimodal, it can understand many types of info. Efficient & flexible, it comes in 3 sizes each best-in-class & optimized for different uses blog.google/technology/ai/…

Excited to share PDGrapher! - a causally-inspired GNN model to predict therapeutically useful perturbagens🧬💊 📜biorxiv.org/content/10.110… 🔬zitniklab.hms.harvard.edu/projects/PDGra… 💫Thank you to amazing collaborators and mentors Marinka Zitnik, Isuru Herath, Kirill Veselkov, Michael Bronstein

Very happy to announce that VeRA is accepted at ICLR 2026 with scores 8,8,8,5! VeRA makes LoRA ~10x more parameter efficient while retaining the same performance & also works for vision! Paper: arxiv.org/abs/2310.11454 Our very light-weight webpage😏: dkopi.github.io/vera/

Our end-to-end workflow for 3D pathology is now published in Nature Protocols! This includes all the steps to go from archived pathology tissues to 3D H&E-like datasets, with an emphasis on quality control for large studies. Full text at: rdcu.be/dwIYW

Text understanding with #LLMs is useful but not enough for scientific understanding and discovery. In chemistry, in addition to text, chemical structure is essential to determine the properties of molecules. We have created the first multimodal text-chemical structure model:

ReacLLaMA: Merging chemical and textual information in chemical reactivity AI models Model uses both molecular structure and textual information on the chemical procedure to predict reactivity. arxiv.org/abs/2401.17267

1/6 🍍🍕🧬 What do a photo of Italians savoring pineapple pizza and a synthetic DNA sequence have in common? Both can be intricately crafted by generative AI! Introducing DNA-Diffusion, our AI model set to advance #SyntheticBiology and #GeneRegulation: biorxiv.org/content/10.110…

A Novel RAG Approach That Understands The Whole Document Context RAG has rapidly evolved to be the standard way to apply LLMs in production. However, most methods are still limited because most existing methods retrieve only short contiguous chunks from a retrieval corpus,

🤩🤩🤩 Spatial Omics fans, I am super excited to share our latest manuscript!!! 🤩🤩🤩 #spatialomics #spatialbiology #singlecell Spatially Exploring RNA Biology in Archival Formalin-Fixed Paraffin-Embedded Tissues biorxiv.org/content/10.110…

New in Nature Methods! We introduce TISSUE, an uncertainty-aware framework for integrating and using #spatial #transcriptomics. Conformal prediction + spatial biology enhances downstream analyses and discovery 💯 Paper: rdcu.be/dyqAj Code: github.com/sunericd/TISSUE 1/3

Announcing Stable Diffusion 3, our most capable text-to-image model, utilizing a diffusion transformer architecture for greatly improved performance in multi-subject prompts, image quality, and spelling abilities. Today, we are opening the waitlist for early preview. This phase

#CellCharter A powerful single-cell #SpatialTranscriptomics analysis framework Graph-based #DeepLearning Spatial cluster: Cell type Different expression Neighborhood enrichment AND Shape!!!👹 - curl/elongation/linearity/purity #VariationalAutoEncoder for dimension reduction,

The latest ML/AI research (Part 1): ▪️ OMNIPRED ▪️ AgentScope ▪️ YOLOv9 ▪️ ChunkAttention ▪️ TinyLLaVA ▪️ OpenCodeInterpreter ... 🧵

Is DNA all you need? Introducing Evo, a long context 7B foundation model for biology Evo has SOTA *zero-shot* prediction across DNA, RNA, and protein modalities Evo can generate DNA, RNA+proteins & make CRISPR-Cas systems for first time blog …n-model-tool-arc-institute.vercel.app/news/blog/evo

⚡️🔬📣Excited to share our two new Nature Medicine articles, we develop computational pathology foundation models, 1. UNI, a self-supervised computational pathology model trained on 100 million pathology images from 100k+ slides. 2. CONCH, a vision-language model for

🤖🧬Techbio deals with biopharma and $NVDA

⚡️🔬📣Excited to share our new Nature Medicine article, examining disparities in pathology AI models, assessing how modeling choices impact disparities, and evaluating the potential of self-supervised foundation models in mitigating these disparities. nature.com/articles/s4159… See

Google announces Med-Gemini, a family of Gemini models fine-tuned for medical tasks! 🔬 Achieves SOTA on 10 of the 14 benchmarks, spanning text, multimodal & long-context applications. Surpasses GPT-4 on all benchmarks! This paper is super exciting, let's dive in ↓

⚡️📣👇Tremendously excited to share our new Cell article, where we develop TriPath, a method for analyzing 3D pathology samples using weakly supervised AI. Article: authors.elsevier.com/a/1j3RiL7PXqQM-. TriPath enables 3D computational pathology via 3D multiple instance learning