Roman Kasianov

The Arc Institute has launched the Virtual Cell Challenge, a new open competition to benchmark computational models that predict how human cells respond to genetic perturbations. Announced today and published in Cell (Vol. 188, Issue 13), the challenge invites AI and computational biology researchers to submit predictive models for scoring on a live leaderboard, using single-cell RNA sequencing data from human embryonic stem cells.

See also: Arc Institute Releases its First Virtual Cell Model

SandboxAQ has launched what it claims is the largest public dataset of protein-ligand complexes with associated binding affinity data. The database, called SAIR (Structurally Augmented IC50 Repository), contains approximately 5.2 million synthetic 3D structures covering over 1 million protein-ligand systems, designed to support the development and benchmarking of AI models in drug discovery.

Context

Spun out of Alphabet in 2022, SandboxAQ operates at the intersection of AI and quantum simulation, with a dedicated biopharma arm known as AQBioSim. Backed by over $300 million in funding (~$5B valuation) and partnerships with AstraZeneca, Sanofi, and the Michael J. Fox Foundation, the company develops Large Quantitative Models (LQMs) to accelerate molecular design and biological simulation.

A new AI model from Google DeepMind, called AlphaGenome, has been released in preview for non-commercial research, offering a high-resolution, long-context DNA sequence model designed to predict the regulatory effects of genetic variants at scale.

XtalPi has signed a letter of intent with DoveTree LLC, a company founded by Harvard University professor Gregory Verdine, to develop small molecule and antibody drug candidates using XtalPi’s AI- and robotics-powered discovery platform. The collaboration will focus on multiple targets selected by DoveTree in oncology, autoimmune disorders, and neurological diseases.

Arc Institute has launched its first virtual cell model, State, designed to predict gene expression changes in response to chemical, genetic, or cytokine perturbations across diverse cell types. The model is publicly released for noncommercial use and trained on observational data from nearly 170 million cells and perturbational data from over 100 million cells across 70 human cell lines, making it the largest single-cell perturbation model to date.

NVIDIA has released a new AI-driven workflow, the Biomedical AI-Q Research Agent Developer Blueprint, designed to accelerate the early phases of drug discovery by automating literature review and hypothesis generation. Built on top of NVIDIA’s AI-Q and Retrieval-Augmented Generation (RAG) blueprints, the system combines a reasoning agent with virtual screening capabilities, targeting use cases such as target identification and small molecule pairing.

San Francisco-based startup Axiom Bio, previously backed by $15 million in seed funding from investors including Amplify Partners, Dimension Capital, and Zetta Ventures, has released a publicly accessible data exploration tool featuring a primary human liver toxicity dataset. This dataset includes detailed liver toxicity profiles covering more than 130,000 distinct molecules.

Xaira Therapeutics has announced the release of "X-Atlas/Orion", a large genome-wide Perturb-seq dataset, reportedly the largest publicly available resource of its kind. The dataset includes information from 8 million individual cells, targeting all human protein-coding genes, with deep sequencing coverage of over 16,000 unique molecular identifiers (UMIs) per cell.

This resource was generated using Xaira's "Fix-Cryopreserve-ScRNAseq" (FiCS) Perturb-seq platform, which leverages the Chromium system from 10x Genomics. According to Xaira, the platform addresses common limitations associated with large-scale single-cell data collection, including sensitivity, scalability, and reproducibility. Xaira’s team indicates that FiCS Perturb-seq reliably captures transcriptional changes resulting from gene perturbations and effectively reflects known biological pathways.

Isomorphic Labs has appointed Dr. Ben B. Wolf as Chief Medical Officer and opened a new U.S. office in Cambridge, Massachusetts, marking a strategic expansion following its $600M raise as the DeepMind spinout strengthens clinical leadership and infrastructure to advance its AI-based drug discovery platform toward the clinic.

AstraZeneca has signed a licensing and co-discovery agreement with CSPC Pharmaceutical Group worth up to $5.3 billion, including $110 million in upfront cash. The partnership will focus on discovering new oral drugs targeting multiple chronic disease indications, including at least one immunological condition.

This follows a prior $100 million deal in October 2024, in which AstraZeneca licensed a CSPC cardiovascular candidate. The new agreement reportedly deepens that collaboration, though specific disease areas or targets under the new work were not disclosed.