Skip to content
HUMAN BIOLOGICAL
SANDBOX

Simulate the body.
Explore how cells change.

Build a research baseline, explore documented biological models, and compare possible cell changes over time.

View the present. Explore the past.

View the present. Explore the future.

See a currently affected cell among moving surrounding cells, then follow an illustrated earlier pathway from a source region to that cell. The demo also explores possible future changes and keeps a traceable example record.

Affected cells

Simulate biological processes to explore how cells could change under different conditions—and examine possible pathways leading to those changes.

View more →
OpenAI knot logoNVIDIACleveland ClinicEMBL-EBIReactomeNYU Langone HealthOHSUOntario Institute for Cancer Research

Begin with the body.
Explore what could change.

The sandbox starts with a baseline biological state and explores alternative trajectories under explicit assumptions. Follow a cell through modeled time to examine a change in its DNA and the questions it raises. The animation illustrates synthetic cell events.

SYNTHETIC CELL 027

DNA change detected

Example event: a single-base substitution.
Sequence: A → G
Compartment: cell nucleus
Next step: compare lineage history.

Hypothetical event, not patient data or a disease prediction.

0Days
0Weeks
0Months
0Years

Build a baseline

Define the starting state, selected biological processes, parameters, and limits of a narrow model.

Branch into futures

Run reproducible ensembles to explore different modeled events over a chosen time horizon.

Follow the change

Track a flagged cellular lineage backward through its simulated history to examine preceding transitions.

Ask a better question

Compare counterfactual scenarios and identify hypotheses that deserve independent experimental research.

One starting body.
Many possible cases.

Start with a person’s information. Explore different conditions, follow possible cell changes, and save cases that need review.

One starting cell and three possible modeled outcomesA blue starting cell connects to three moving cells. One stays blue, one briefly changes and returns to blue, and one turns red for review.

One starting cell. Three possible cases: the top cell stays unchanged; the middle cell changes and recovers; the bottom cell keeps its change and turns red for review. The engine explores cases from a person’s starting information over five years. This is an illustration of the workflow.

How cells are produced

Explore how real cells divide, how digital cells represent them, and how published evidence helps interpret DNA changes.

Explore cell production

The future properties can break the past

Begin with a known cell-change record. Explore possible future pathways from candidate starting properties, then compare their outcomes with that record. A matching pathway offers a possible explanation of how the given state could arise.

White, pink and blue loops represent alternative cases. A red point marks a match: movement stops, the loops turn red, and the model record opens.

View the reference record →
Three layers of infinity pathways finding a model matchThin white, pink and blue infinity lines and moving dots compare synthetic cases. A matched point turns red, all motion stops, the lines turn red and a candidate explanation appears.
Exploring candidate cases0 / 3 compared
Three synthetic pathways · one reference state

Latest research records

Show more →

Research Forum

Share research questions, discuss biological models, and connect with other members. Read public discussions freely; sign in to post, reply, or send private messages.

Our mission

Start with your measurements, explore possible changes, and keep the results for review. Build understanding one model at a time.

Explore the process

The three-component plan

  1. Collect · Build the digital body. Create a starting database from the person’s measurements.
  2. Run properties · Vary sizes. Give each case its own supported sizes and values.
  3. Age · Follow size changes. Version 1 is planned to explore property-size evolution over five years.

Here, “aging” describes property-size evolution. Person age remains a separate input. This engine plan is proposed.

Explore Engine Version One

Measure & model

Enter age, weight, and height. Review the supported estimates and add any missing measurements.

Explore & record

Explore possible changes. Keep each review case with its inputs, calculations, and sources.

Validate & expand

Check the model against separate measurements and expert review. Then extend the sandbox to more cell types, tissues, and biological processes.

Baseline → futures

Start with a biological model

Trace the lineage

Follow the modeled sequence of change

Test hypotheses

Compare interventions inside the sandbox

Research articles

What we learned building the population engine: successful checks, failed attempts, and what comes next.

Read the overnight review →All articles →

The idea is documented.
The work is beginning.

Read the project foundations, examine the assumptions, and explore what comes next.

AI-assisted analysis and centralized simulation tickets are proposed architecture components. Selected research models can run in the workspace; their results depend on the supplied inputs and model scope.