computation · physical biology
The virtual cellcompany.
Biology can describe a cell in extraordinary detail. It still can’t tell you what happens if you change one gene.
The biological choices that shape a therapy should be predictable before a patient depends on them.
That is the question Galen answers: causal models of living cells that predict what changes under intervention before the next experiment begins.
why Galen exists
The most consequential choices in medicine begin upstream.
Long before a therapy reaches the clinic, scientists decide which biological changes are worth testing. Those choices determine where years of work and scarce experimental capacity are spent.
We start with a problem where the choice is everything: engineered T cells that work, then stop working. Durability and exhaustion in engineered cell therapies is a question about what a living system does next — exactly the question a descriptive model cannot answer.
Galen exists to make those choices predictable. Our mission is to solve disease from first principles.
A model of biology must remain answerable to reality.
why now
Biology can now be measured at extraordinary resolution. The frontier is knowing what will change — and choosing it.
Biology has maps of cellular state and laboratories that can test reality. What is missing sits between them: the models that read those maps describe what cells look like, not what they do when something is changed.
That gap is causal. Correlation can tell you which measurements travel together; only a causal model tells you which intervention moves the system, and how far.
Living systems are adaptive, context-dependent, and difficult to control. A useful virtual cell must therefore stay close to experiment: prediction first, measurement always, learning next.
Galen starts at the cell because that is where biological change becomes measurable enough to predict — and then to design with.
the loop
Make the next experiment more deliberate.
A virtual cell should not replace experiment. It should make the next experiment more deliberate by turning cellular state into a claim that can be measured, corrected, and used to ask a sharper question.
- 01
Predict
State what changes before the next experiment begins.
- 02
Measure
Let physical biology decide what the model got right.
- 03
Learn
Use each result to sharpen the next question.
- 04
Design
Turn measured understanding into deliberate biological work.
what it changes
Spend physical biology on stronger questions.
The cost of biological work is not only the price of an experiment. It is the time spent testing ideas before the system has made clear which ideas deserve reality next.
Galen is building a different order of operations: ask the question in computation, commit the strongest claims to experiment, and let the result improve the next design.
Ask the question in computation. Spend reality on the strongest claims.
- Less blind search
- Use models to compare possible changes before committing lab time, budget, and cells.
- Better uncertainty
- Make ignorance visible so the next measurement can reduce it.
- More deliberate design
- Move from observing life toward designing with living systems, one measured claim at a time.
work with us
A frontier worth building.
If you are building, funding, or partnering around virtual cells, programmable biology, scientific software, or the interface of computation and living systems, we want to hear from you.
Galen is for scientists, engineers, partners, funders, and long-term company builders who believe biology should be predictable, measurable, and deliberately designed.