Generative model designs working disordered protein segments from scratch
A biophysical generative algorithm designed functional intrinsically disordered regions from scratch for the developmental transcription factor RUNX2.
Benchmark tests whether AI models actually use biological evidence to predict cell perturbations
Researchers built CellChain, an evaluation benchmark spanning 590 biological cases and 2,360 matched conditions across gene dependency, cytokine signaling, and drug sensitivity.
Soft thermodynamic penalties keep biochemical pathway models physically sound
Researchers introduced an optimization and Bayesian inference framework that penalizes parameter sets violating thermodynamic detailed balance, rather than strictly reparameterizing complex reaction networks.
Single-cell pipeline infers gene regulatory networks and cell types simultaneously
Computational biologists introduced NetMap, a software pipeline that jointly infers single-cell gene regulatory networks and clusters cell populations without needing pre-assigned cell annotations.
Vision language models interpret multiplexed cell imaging screens
Genentech researchers developed CellMAPP and an analytical framework called CHART to track organelle changes across 524 genetic knockdowns in breast cancer cells.
Generative model designs culture conditions to steer human blood stem cells
Researchers built LabCompass, an active-learning system that pairs a flow-matching generative model with a surrogate predictor to design continuous chemical recipes that push stem cells toward specific fates.
Physical cell crowding guides reinforcement learning for adaptive cancer therapy
Researchers combined two-dimensional melanoma cell cultures with mechanical particle simulations to study how spatial crowding drives drug-resistant tumor cells to break free.
Language model predicts molecular dynamics trajectories across diverse proteins
Researchers introduced MD-LLM-2, a transferable language model that generates protein motion trajectories conditioned on sequence, temperature, and structural history.
Diffusion model predicts unseen genetic perturbation effects and maps causal gene networks
Researchers introduced latent causal diffusion (LCD), a generative diffusion framework designed to model single-cell perturbation responses while accounting for measurement noise.
Chan Zuckerberg Biohub launches $1.8 billion AI virtual cell partnership
The Chan Zuckerberg Biohub has unveiled an expanded $1.8 billion Virtual Biology Initiative in collaboration with Meta, Google DeepMind, Isomorphic Labs, the Department of Energy, and the National Institutes of Health.
CNBC, also reported by Endpoints News (subscription required)Yahoo FinanceOptimal transport model infers gene networks and missing protein levels from single-cell snapshots
CardamomOT combines optimal transport mathematics with a mechanistic model of gene expression to infer gene regulatory networks and reconstruct unmeasured protein trajectories across single-cell RNA sequencing time series.
Tool turns single-cell data into executable gene-network models
A computational framework called scBOLT simplifies the process of turning single-cell RNA sequencing data across multiple experimental conditions into Boolean logic models of gene regulation.
Generative flow model steers single-cell predictions across new perturbations
Researchers introduced scEGFlow, a framework combining conditional flow matching with energy guidance to model how individual cells shift in appearance and gene expression following biological perturbations.
Stand-alone software models gene networks to find master regulators in single-cell data
Computational biologists developed SCITRAM, a desktop tool that infers gene regulatory networks from single-cell RNA sequencing and simulates transcription factor activation cascades to identify master control proteins.
Benchmark tests how well AI interprets biology figures for virtual cells
OmniVCBench introduces 6,077 question-and-answer pairs based on published figures to evaluate whether multimodal AI models can interpret wet-lab data and formulate biological hypotheses.
Generative model maps branching cell trajectories from static RNA snapshots
KoopCell uses dynamical systems theory to infer cell development over time from unpaired single-cell RNA sequencing snapshots.
Autonomous AI agent derives biological equations and mechanisms directly from data
A team created GEMOT, an AI agent framework designed to infer explicit, modular mathematical equations and hidden biological states from sparse experimental data.
Generative model maps causal effects of gene edits and drugs in single cells
Researchers introduced SHERLOCK, a deep generative model that treats single-cell genetic and drug perturbations as structured interventions on a baseline cellular state within a causal framework.
Biophysical simulations find alternative ion channel targets to balance disease-causing mutations
An automated modeling pipeline uses biophysical simulations of neurons to identify how adjusting normal ion channels can restore healthy firing patterns without directly targeting the mutated channel.
Deep learning links genome sequence models to single-cell data to map gene networks
DeepSCENIC transfers representations from sequence-to-function deep learning models, such as Enformer and Borzoi, to single-cell multiomic atlases to reconstruct causal gene regulatory networks.
AI infers tissue mechanical stress directly from cell shapes
StressNET, a graph neural network, estimates physical forces between neighboring cells directly from microscopy images without relying on pre-specified physical formulas.
Projecting single-cell expression data onto a sphere yields interpretable developmental coordinates
Computational biologists developed SPHERE-PCA, a method that maps the first three principal components of single-cell RNA sequencing data onto a three-dimensional unit sphere to create an interpretable coordinate framework.
Generative model translates protein sequence into microscopy images
Researchers built CELL-FM, a bidirectional generative AI framework that predicts fluorescence microscopy images directly from protein sequences and cellular context, and can also work in reverse.
Digital twin model simulates reduced-dose epigenetic drug schedules in breast cancer
Researchers built a mechanistic computational model combining chromatin dynamics, in vitro response data, and drug distribution kinetics to simulate how triple-negative breast cancer responds to combination epigenetic therapy.
Therna releases synthetic gene atlas measuring 60,000 regulatory elements
Therna Biosciences has released public datasets from its Chronos platform profiling roughly 60,000 synthetic regulatory elements across approximately 50 pooled cell lines.
AI pipeline chains single-cell models to screen drug combinations
Computational biologists introduced PHAROS, an algorithmic framework that chains predictions from existing single-cell perturbation models to simulate and search for synergistic two-drug combinations.
Benchmark reveals conservative bias and cell-type limits across perturbation models
A systematic evaluation of 13 computational methods across 25 perturbation datasets found that current models struggle to generalize cellular drug and gene responses across different cell types.
Virtual cell model maps dynamic protein shifts across 38 million measurements to predict drug response
Scientists developed ProteinTalks, an operational virtual cell model pretrained on more than 38 million time-resolved protein abundance measurements from perturbed breast cancer cell lines.
Neural network decodes developmental signaling histories in living tissues
The computational tool IRIS uses transfer learning to infer in vivo signaling dynamics and cellular histories from single-cell gene expression atlases.