Data assessment and study planning
Discuss data suitability, task boundaries and analysis plans using sample provenance, controls, batches, usage rights and study endpoints.
AI FOR CELLULAR DISCOVERY
Understand cell states
Explore drug responses
Connect single-cell data with AI to explore how cells change under different interventions and define clearer directions for experimental validation.
CELL STATE · PERTURBATION · DRUG RESPONSE
CELLULAR INTELLIGENCE
From observation to prediction.
From observation to prediction.
IN SILICO / CELLULAR LANDSCAPE
Conceptual cell visualization · Not microscopy
Conceptual cell visualization · Not microscopy
RESEARCH DIRECTIONSExplore key questions at the cellular level
01 / STATECell state prediction
02 / RESPONSEDrug response simulation
03 / PERTURBATIONGenetic and environmental perturbations
PROJECT SERVICES
Technical services for cellular research
Discuss the scope around your data and research question. Work packages and acceptance criteria are confirmed in the project agreement.
Cell state and perturbation analysis
Define cell populations, differential responses, candidate pathways and model selection around project data. Separate associations, predictions and mechanistic hypotheses.
Model evaluation and validation planning
Agree on held-out tests, comparison baselines, evaluation metrics and experimental follow-up conditions for reviewable project outputs.
RESEARCH FOCUS
Turn complex cellular questions
into focused computational tasks.
Connect data, models and experiments through clearly defined questions about cellular state and response.
Cell state prediction
Characterize cellular heterogeneity and state transitions to explore differentiation, stress and disease-associated changes.
Learn more ↗Drug response simulation
Investigate cellular responses to candidate compounds in the context of cell background, dose and exposure time.
Learn more ↗Perturbation analysis
Explore expression and pathway responses to genetic or environmental changes, generating experimentally testable hypotheses.
Learn more ↗FROM DATA TO EVIDENCE
Connect models and experiments
around one research question.
A useful prediction helps researchers decide what to validate next.
Define the question
Specify cell types, interventions and endpoints to establish the scope of the task.
Organize the evidence
Connect controls, doses, time points and batch metadata into a consistent data view.
Frame the prediction task
Choose appropriate representations and metrics, compare baselines and assess uncertainty.
Design validation
Translate candidate rankings and mechanistic hypotheses into experiments, then evaluate the results.
APPLICATIONS
Start with the research objective.
Drug discovery and candidate prioritization
Compare cellular responses across intervention conditions to guide candidate prioritization and follow-up experiments.
↗02Cell differentiation and state changes
Analyze heterogeneity and state trajectories across culture and intervention conditions.
↗03Disease mechanisms and target exploration
Connect perturbation signals to pathway changes and develop testable mechanistic hypotheses.
↗INSIGHTS
Understand the methods and their limits.
What is an AI virtual cell? From representation to perturbation prediction
Understand the scope, core tasks and limitations of virtual cell research.
Read article ↗Why dose and time matter in drug response simulation
Consider compounds, cellular background and experimental conditions within one framework.
Read article ↗How should cell prediction models be validated?
Build a traceable evidence chain through data separation, baseline comparisons and external validation.
Read article ↗QUESTIONS & ANSWERS
About AI virtual cells
What is an AI virtual cell?
An AI virtual cell uses data and computational models to represent cellular states and study responses to drugs, genetic changes or environmental conditions. Individual models cover selected cell types, modalities and interventions; they are not complete replicas of living cells.
How do cell state prediction and drug response simulation differ?
Cell state prediction concerns current or future molecular, morphological and functional features. Drug response simulation additionally considers compounds, doses and exposure time to study changes following treatment. They may share representations but require different tasks and evaluation criteria.
What data are needed to get started?
Define the task first. Single-cell transcriptomic tasks typically require expression matrices, cell and sample metadata, controls, perturbation conditions, doses and time points. Imaging tasks require source images, acquisition settings and relevant labels. Suitability depends on data quality, coverage and usage rights.
START WITH A QUESTION
Begin with a cellular question.
Clarify the objective, data and validation pathway so predictions can inform the next experiment.