Eric Minwei Liu @EricM_Liu

Human iPSC-to-motor-neuron differentiation shows a transient mtDNA drop, then recovery with maturation. Temporal omics map pyruvate rewiring, chiefly in one iPSC line with technical well replicates. biorxiv.org/content/10.648… #StemCells #bioRxiv

We now have state-of-the-art pre-trained base models in the 3T-parameter weight class, built entirely in-house at MAI. Post-training results coming soon -- stay tuned! Lots of details in our technical blog post. microsoft.ai/news/frontier-…

Can targeting PKMYT1 overcome cisplatin resistance in #Osteosarcoma? This study links PKMYT1-mediated NPM1 phosphorylation to altered DNA repair factor recruitment and shows that RP6306 enhances cisplatin #Cytotoxicity. #STTT #OpenAccess: doi.org/10.1038/s41392…

Review @jclinicalinvest Selective estrogen receptor degraders and their effect on the breast cancer immune microenvironment jci.org/articles/view/…

Beyond the obvious pocket: nature.com/articles/s4158…

In the latest issue! Metabolic atlas of early human cortex reveals glycolytic remodeling and pentose phosphate pathway control of cell fate transitions dlvr.​it/TVyZNj

Robotics has spent years scaling demonstration data and model size. What happens when we scale EXPLORATION for robot agents? We're incredibly excited to share RAX, an agentic robotics system that learns by scaling online exploration. A coding agent connects directly to the robot’s perception and control APIs, learns from feedback, and builds programs that perceive, plan, and act. The model stays frozen. As exploration scales, the agent improves its programs and builds up reusable skills. The results: • ZERO teleoperation or action demonstrations on real robots—from insertion with 0.1 mm clearance and air-fryer manipulation to deformable objects and long-horizon mobile whole-body control. • +37 percentage points on average over leading learned policies across five simulation benchmarks: RoboDojo, BEHAVIOR-1K, HumanoidArena, DexJoCo’s bimanual tasks, and LIBERO-Pro. • One framework spanning single-arm, bimanual, dexterous, humanoid, and mobile robots, with new embodiments connected through APIs. The craziest part: on each real-world embodiment, going from scratch to mastering the tasks took roughly two hours. No teleop. No model finetuning. We’re open-sourcing the full framework AND every experiment session. We can’t wait to see what people build with it! Project + sessions: robot-agent-exploration.com Paper: assets.robot-agent-exploration.com/paper/rax.pdf GitHub: N Project fully led by @RunhanH and @xczhangxx who are absolutely awesome. How it works, experiments, and robot videos below 🧵 (1/N)

scBaseCount: An AI agent-curated, standardized, auto-updated single-cell data repository Preprint link biorxiv.org/content/10.110… cell.com/cell/fulltext/… 1. A massive single-cell gene expression resource: scBaseCount is presented as the largest public repository of uniformly processed 10x Genomics single-cell RNA-seq data, containing over 502 million cells across 27 organisms and 75 tissues. 2. Addresses underutilized public data: The Sequence Read Archive (SRA) contains enormous amounts of sequencing data, but inconsistent metadata makes datasets difficult to discover and compare. scBaseCount aims to unlock this information through automated discovery and standardization. 3. AI agents automate data curation: The researchers developed SRAgent, an AI-powered, hierarchical workflow using large language models (LLMs) and specialized tools to identify datasets, retrieve metadata, and annotate tissues, organisms, and diseases. 4. Substantial improvements in scale and efficiency: SRAgent processed 208,939 SRA experiments and identified 105,343 as 10x Genomics libraries. The study estimates that its automated approach saved more than 500,000 CPU hours compared with brute-force dataset identification. 5. Standardized processing improves comparability: The team developed scRecounter, a scalable Nextflow pipeline that uniformly reprocesses sequencing data using STARsolo. It supports different gene-counting strategies and multi-mapping approaches, reducing analytical inconsistencies across studies. 6. Expanded representation of biological diversity: Compared with resources such as CZ CELLxGENE and the Human Cell Atlas, scBaseCount includes a broader range of organisms, tissues, and experimental contexts. The collection also reveals trends in sequencing technologies, disease studies, and tissue representation. 7. Enables cell-type annotation and cross-species analysis: The researchers applied computational models and reference atlases to annotate cells from humans, mice, zebrafish, and Drosophila. Their analyses identified shared cell populations and generated resources for comparing conserved and divergent cell-type signatures across species. 8. Provides useful non-coding and intronic expression data: The standardized pipeline supports multiple counting strategies, including intronic and exonic reads and non-coding gene expression. These features expand the resource's utility for studying transcriptional dynamics and other aspects of cellular biology. 9. Supports AI foundation models for biology: The study reports that training a cell-embedding model on scBaseCount improves its ability to distinguish diverse cellular phenotypes. The scale and diversity of the resource could help develop models that predict cell states in previously unobserved biological contexts. 10. Establishes a model for autonomous scientific data infrastructure: By combining AI agents, standardized pipelines, and continuous dataset discovery, scBaseCount provides a framework for maintaining large biological databases with less manual curation. The authors have also made SRAgent and scRecounter code available to support further research and development. Key takeaway: scBaseCount illustrates how AI can transform scattered public sequencing datasets into a large, standardized resource for biological discovery and foundation-model development. Its major contribution is not simply the number of cells collected, but the combination of automated curation, uniform processing, cross-species diversity, and AI-ready data.

🧠 Yann LeCun helped invent the convolutional neural network and shared the 2018 Turing Award with Hinton and Bengio. Now he joins us at AI Dev NYC to share his latest work at AMI Labs. 📍 NYC 🗓️ Nov 30 to Dec 1 👉 ai-dev.deeplearning.ai Just. Keep. Building.

Transcriptomic reprogramming screen identifies SRSF1 as a regulator of aging: nature.com/articles/s4358…

Detecting low-allele-fraction mosaic mutations: A multi-institution benchmarking with cell-line mixtures dlvr.​it/TVy8QS

In the latest issue! CRISPR screens identify targets to rescue age-related T cell dysfunction in cancer dlvr.​it/TVy83b

A phase 1 trial shows that treatment of patients with high-grade serous ovarian cancer with personalized neoantigen vaccines combined with PD-1 blockade was feasible and led to tumor-specific immune responses in many patients. dlvr.​it/TVwzPY @DFCI_NeoVax

How tissue dissociation impacts scRNA-seq: Standard protocols can lose reactive astrocytes & key pathological signatures! 🧠 Check out our new paper in Molecular Neurobiology on optimized workflows: link.springer.com/article/10.100… #SingleCell #Neuroscience #Astrocytes

A pangenomic method for establishing a synthetic somatic variant resource in HapMap mixtures dlvr.​it/TVy58y

Here the 50+ ADCs that will be discussed at #ESMO26 next week, divided by payload class and target. B7-H3 TOPO1i leads, with 7 different ADCs! (Note: TiP and ePoster not included)

Raffaele Colombo @raffcolo

a week ago

Only 2 weeks until the @myESMO 2026 Congress in Madrid! For ADCs: ▪️ 210+ regular abstracts and 23 Late-Breaking Abstracts (LBAs) accepted ▪️ 50+ talks, 120+ posters, and 60+ ePosters 👇👇 Here the list of the 14 ADCs being presented for the first time at #ESMO26!

2 24 60 21K 30

Looks cool to me 😮 198 simulation-ready chemistry laboratory instruments. Each one comes with part-level geometry, working joints, an executable behaviour model and a contract that records what every value means and where it came from. All 198 instruments have been converted to USD and stepped in NVIDIA Isaac Lab through the Matterix framework, with joints driven from runtime state and panel buttons pressed in the scene reaching the runtime. mbzuai-nlp.github.io/ChemLab-Assets/ huggingface.co/datasets/MBZUA…

here is an overview of the architecture of all the major american open weight models lots of similarities: sliding window (except mamba for nvidia), latent moe or a variant from muse, and a LOT of norms everywhere (muse has 9 norms per layer if i counted right!!) let's focus on muse, it's the only model in this list without shared experts. they are also doing different residuals with some form of gate (still single stream), and are more conservative in the hybrid ratio (3:1) and larger sliding window (2k). there are not a lot of details on the training procedure but you can notice there are a bunch of "muP like" multipliers including one for the router with a 1/sqrt(nb of experts) factor 👀 super interesting to see this quite different approach to scaling between american and chinese open weight labs, different constraints lead to different architectures

Alexandr Wang @alexandr_wang

3 days ago

1/ muse spark open is the most capable US open-weight model to date. 816B total params, 42B active. 1M-token context. natively multimodal. built for coding and agentic work like long-horizon tasks, hard software engineering, multi-step reasoning. america! 🇺🇲

294 462 6K 667K 748

Why do some breast cancer cells lie dormant before suddenly spreading? A genetic screen identified MED4 as a key gatekeeper ➡️ loss of one copy drives enhancer reprogramming and metastatic outgrowth, pointing to a new predictive biomarker. nature.com/articles/s4155…

What Do Oral Drugs Really Look Like? Dose Regimen, Pharmacokinetics, and Safety of Recently Approved Small-Molecule Oral Drugs Dean G. Brown * J. Med. Chem. (2025) 68 (22): 23751–23780. pubs.acs.org/jmcmar/article…