From biological trial and error to guided discovery
Modeling cells
Better treatment candidates, more precise experiments, and eventually predictive models of how living cells respond.
The unlock
AI combines molecular structures, gene activity, and experimental perturbations to propose what might happen when a cell is changed.
The hard part
A protein structure is not a whole cell, and a cell is not a whole patient. Predictions require laboratory validation and treatments require clinical evidence. Cure scenarios below are research possibilities, not available cures.
How to read these scenarios
These are positive possibilities, not promises or timeline forecasts. Lists put the most intriguing ideas first: an editorial judgment about interest and transformative potential, not likelihood or arrival date. The labels describe what must happen next, not numerical probabilities.
Already emerging: early versions or important components exist; the full scenario may still need work.
Plausible next step: an extension of demonstrated capabilities, with meaningful engineering and adoption work remaining.
Breakthrough-dependent: needs a scientific or technical advance that is not assured.
Products are useful tools and services; cures include potential treatments and prevention, not claims of proven cures; insights are discoveries that help us understand the world.
Better control of cell identity could help repair damaged cartilage or other tissues in selected conditions.
Regeneration requires more than producing the right cell type. Cells must form an appropriate structure, interact with their surroundings, and remain functional over time. Modeling could help researchers choose signals, scaffolds, and conditions that support that process.
Some tissues are especially difficult to restore because their architecture, mechanical properties, or blood supply are complex. A laboratory culture that resembles healthy tissue is an encouraging intermediate result, not proof of a durable repair in a person.
What this could look like
Researchers combine cell-state predictions with scaffold experiments to seek a cartilage repair that integrates with surrounding tissue and withstands repeated loading.
Who benefits, and how
Better-directed experiments in tissue repair.
Potential restorative treatments for selected injuries or degenerative conditions.
What would make it real
Show stable function and integration in relevant models, not just favorable cell markers.
Establish durability and rule out unacceptable immune responses or uncontrolled growth.
Organ-level structure, blood supply, immune compatibility, durability, and cancer risk are not solved by cell-state prediction.
Some cancers become more curable through combinations that attack a tumor's vulnerabilities while sparing healthy tissue.
Why AI helps
Validated cell models could help identify resistance pathways and prioritize combination therapies.
What still has to happen
Tumor evolution, immune effects, drug delivery, toxicity, and clinical trials remain essential. This is not a universal cancer cure.
Read details ↗Full details: More targeted cancer treatments
Modeling cells · Cures & treatments
More targeted cancer treatments
Breakthrough-dependent
Some cancers become more curable through combinations that attack a tumor's vulnerabilities while sparing healthy tissue.
A tumor is not a uniform population of cells. Different clones can respond differently to treatment, and resistant populations can become dominant. Predictive models might help identify vulnerabilities shared across those populations or combinations that reduce the chance of escape.
A promising combination must also spare enough healthy tissue to be tolerable. Understanding a cancer cell in isolation is insufficient: immune interactions, tissue context, and the distribution of drugs through the body affect whether a treatment succeeds.
What this could look like
For a particular cancer subtype, researchers investigate whether blocking two complementary pathways suppresses both the main tumor population and a resistant subpopulation. The model proposes the hypothesis; experiments and trials determine its value.
Who benefits, and how
More rationally selected combinations for specific cancers.
Potentially better outcomes for patients whose tumors resist existing options.
What would make it real
Show a meaningful therapeutic window between tumor effects and damage to healthy tissue.
Demonstrate durable patient benefit in appropriately designed clinical trials.
Tumor evolution, immune effects, drug delivery, toxicity, and clinical trials remain essential. This is not a universal cancer cure.
Small patient populations gain more candidate therapies, including for some disorders caused by a single gene.
Why AI helps
Models could connect a genetic change to disrupted cellular processes and propose corrections or repurposed drugs.
What still has to happen
Delivery to the right tissue, off-target effects, long-term safety, and trial feasibility still constrain progress.
Read details ↗Full details: Treatments for neglected rare diseases
Modeling cells · Cures & treatments
Treatments for neglected rare diseases
Breakthrough-dependent
Small patient populations gain more candidate therapies, including for some disorders caused by a single gene.
For some rare disorders, a known genetic change offers a starting point, but the steps from that change to illness are poorly understood. Models could help identify which cellular functions are disrupted and whether restoring a function might help.
Possible interventions include existing drugs, new molecules, or genetic approaches, depending on the disorder. None follows automatically from identifying the mutation. Reaching the affected cells and intervening at the right stage of disease can be the hardest problems.
What this could look like
A team studying a rare enzyme deficiency uses patient-derived cells and computational models to investigate whether an existing compound can restore a measurable function. A reproducible laboratory result supports further development, not immediate prescribing.
Who benefits, and how
More research hypotheses for conditions with small patient populations.
Better disease models and measurements for evaluating candidate therapies.
What would make it real
Connect the proposed cellular correction to a clinically meaningful outcome.
Establish delivery, dosing, long-term safety, and a feasible evidence pathway for the patient population.
Delivery to the right tissue, off-target effects, long-term safety, and trial feasibility still constrain progress.
Biologists distinguish changes that drive disease from changes that merely accompany it.
Why AI helps
AI can connect large datasets and choose informative perturbation experiments.
What still has to happen
Causal claims need controlled interventions and replication; observational correlations alone are insufficient.
Read details ↗Full details: A clearer map of what causes disease
Modeling cells · Insight
A clearer map of what causes disease
Plausible next step
Biologists distinguish changes that drive disease from changes that merely accompany it.
Large biological datasets reveal many differences between healthy and diseased samples. Some differences drive disease; others are responses to it, or reflections of unrelated differences between the people sampled.
AI can help prioritize which relationships to investigate and which interventions would distinguish competing explanations. A stronger causal claim comes from changing the suspected driver and observing the predicted consequences under appropriate controls.
What this could look like
If a gene is unusually active in diseased cells, researchers perturb it and measure whether the disease-relevant function changes. They also test an alternative explanation, such as inflammation causing both observations.
Who benefits, and how
More defensible treatment targets.
Fewer resources spent on correlations that do not survive intervention.
What would make it real
Design experiments that distinguish the main hypothesis from plausible alternatives.
Replicate the result across relevant cells, contexts, and independent samples.
Causal claims need controlled interventions and replication; observational correlations alone are insufficient.
Researchers screen candidate interventions computationally before spending scarce laboratory time on the most promising ones.
Why AI helps
Models can combine molecular interactions and cellular measurements to prioritize experiments.
What still has to happen
Reliable prediction for unfamiliar compounds, cell types, and patient populations is still difficult.
Read details ↗Full details: A virtual first pass for drug discovery
Modeling cells · Product
A virtual first pass for drug discovery
Plausible next step
Researchers screen candidate interventions computationally before spending scarce laboratory time on the most promising ones.
A useful virtual-cell model would predict selected responses to a proposed intervention: changes in gene activity, cell survival, or a disease-relevant function. It need not simulate every molecule in a cell to help researchers choose a more informative next experiment.
Predictions are most useful when they distinguish promising candidates from uninformative ones and expose uncertainty. The process remains a loop: predict, test in real biological systems, compare the result, and improve the model.
What this could look like
Researchers studying a disease pathway use a model to rank candidate compounds and doses. They test a carefully chosen subset, including controls and candidates the model is uncertain about, before deciding what to study next.
The experiment stays in the loop
Predict
Estimate a specific cellular response and its uncertainty.
Choose
Prioritize informative interventions and controls.
Test
Measure what actually happens in a biological experiment.
Update
Use agreements and failures to revise the model.
↶ Feed the results back into the next round of predictions and experiments.
A conceptual discovery workflow—not a replacement for laboratory studies or clinical trials. Each experiment can change the next prediction.
Who benefits, and how
Better use of limited samples and laboratory capacity.
Earlier identification of weak hypotheses and useful follow-up experiments.
What would make it real
Outperform straightforward baselines on previously unseen interventions.
Reproduce the relevant biological effect in independent experiments and appropriate disease models.
Reliable prediction for unfamiliar compounds, cell types, and patient populations is still difficult.
Better control of cell identity could help repair damaged cartilage or other tissues in selected conditions.
Regeneration requires more than producing the right cell type. Cells must form an appropriate structure, interact with their surroundings, and remain functional over time. Modeling could help researchers choose signals, scaffolds, and conditions that support that process.
Some tissues are especially difficult to restore because their architecture, mechanical properties, or blood supply are complex. A laboratory culture that resembles healthy tissue is an encouraging intermediate result, not proof of a durable repair in a person.
What this could look like
Researchers combine cell-state predictions with scaffold experiments to seek a cartilage repair that integrates with surrounding tissue and withstands repeated loading.
Who benefits, and how
Better-directed experiments in tissue repair.
Potential restorative treatments for selected injuries or degenerative conditions.
What would make it real
Show stable function and integration in relevant models, not just favorable cell markers.
Establish durability and rule out unacceptable immune responses or uncontrolled growth.
Organ-level structure, blood supply, immune compatibility, durability, and cancer risk are not solved by cell-state prediction.
Some cancers become more curable through combinations that attack a tumor's vulnerabilities while sparing healthy tissue.
A tumor is not a uniform population of cells. Different clones can respond differently to treatment, and resistant populations can become dominant. Predictive models might help identify vulnerabilities shared across those populations or combinations that reduce the chance of escape.
A promising combination must also spare enough healthy tissue to be tolerable. Understanding a cancer cell in isolation is insufficient: immune interactions, tissue context, and the distribution of drugs through the body affect whether a treatment succeeds.
What this could look like
For a particular cancer subtype, researchers investigate whether blocking two complementary pathways suppresses both the main tumor population and a resistant subpopulation. The model proposes the hypothesis; experiments and trials determine its value.
Who benefits, and how
More rationally selected combinations for specific cancers.
Potentially better outcomes for patients whose tumors resist existing options.
What would make it real
Show a meaningful therapeutic window between tumor effects and damage to healthy tissue.
Demonstrate durable patient benefit in appropriately designed clinical trials.
Tumor evolution, immune effects, drug delivery, toxicity, and clinical trials remain essential. This is not a universal cancer cure.
Small patient populations gain more candidate therapies, including for some disorders caused by a single gene.
For some rare disorders, a known genetic change offers a starting point, but the steps from that change to illness are poorly understood. Models could help identify which cellular functions are disrupted and whether restoring a function might help.
Possible interventions include existing drugs, new molecules, or genetic approaches, depending on the disorder. None follows automatically from identifying the mutation. Reaching the affected cells and intervening at the right stage of disease can be the hardest problems.
What this could look like
A team studying a rare enzyme deficiency uses patient-derived cells and computational models to investigate whether an existing compound can restore a measurable function. A reproducible laboratory result supports further development, not immediate prescribing.
Who benefits, and how
More research hypotheses for conditions with small patient populations.
Better disease models and measurements for evaluating candidate therapies.
What would make it real
Connect the proposed cellular correction to a clinically meaningful outcome.
Establish delivery, dosing, long-term safety, and a feasible evidence pathway for the patient population.
Delivery to the right tissue, off-target effects, long-term safety, and trial feasibility still constrain progress.
Biologists distinguish changes that drive disease from changes that merely accompany it.
Large biological datasets reveal many differences between healthy and diseased samples. Some differences drive disease; others are responses to it, or reflections of unrelated differences between the people sampled.
AI can help prioritize which relationships to investigate and which interventions would distinguish competing explanations. A stronger causal claim comes from changing the suspected driver and observing the predicted consequences under appropriate controls.
What this could look like
If a gene is unusually active in diseased cells, researchers perturb it and measure whether the disease-relevant function changes. They also test an alternative explanation, such as inflammation causing both observations.
Who benefits, and how
More defensible treatment targets.
Fewer resources spent on correlations that do not survive intervention.
What would make it real
Design experiments that distinguish the main hypothesis from plausible alternatives.
Replicate the result across relevant cells, contexts, and independent samples.
Causal claims need controlled interventions and replication; observational correlations alone are insufficient.
Researchers screen candidate interventions computationally before spending scarce laboratory time on the most promising ones.
A useful virtual-cell model would predict selected responses to a proposed intervention: changes in gene activity, cell survival, or a disease-relevant function. It need not simulate every molecule in a cell to help researchers choose a more informative next experiment.
Predictions are most useful when they distinguish promising candidates from uninformative ones and expose uncertainty. The process remains a loop: predict, test in real biological systems, compare the result, and improve the model.
What this could look like
Researchers studying a disease pathway use a model to rank candidate compounds and doses. They test a carefully chosen subset, including controls and candidates the model is uncertain about, before deciding what to study next.
The experiment stays in the loop
Predict
Estimate a specific cellular response and its uncertainty.
Choose
Prioritize informative interventions and controls.
Test
Measure what actually happens in a biological experiment.
Update
Use agreements and failures to revise the model.
↶ Feed the results back into the next round of predictions and experiments.
A conceptual discovery workflow—not a replacement for laboratory studies or clinical trials. Each experiment can change the next prediction.
Who benefits, and how
Better use of limited samples and laboratory capacity.
Earlier identification of weak hypotheses and useful follow-up experiments.
What would make it real
Outperform straightforward baselines on previously unseen interventions.
Reproduce the relevant biological effect in independent experiments and appropriate disease models.
Reliable prediction for unfamiliar compounds, cell types, and patient populations is still difficult.