We treat symptoms, not causes.
Depression has at least four distinct molecular subtypes. First-line treatment fails 30 to 60% of people. Autoimmune diagnosis takes 4.5 to 7 years and four or more doctors.
The Cytognosis Platform
The GPS for human health.
Medicine waits for damage. We map your biology, track where you are, and help you change course years before symptoms appear.
GPS for human health
Reactive, today
“You have diabetes.”
Proactive, with Cytognosis
“Your metabolic coordinates are shifting. Adjust your circadian timing to restore your trajectory.”
The problem
The same label can hide different biology.
We measure once a year. Biology changes every minute.
Alzheimer’s begins decades before memory loss. Type 2 diabetes is 50 to 80% reversible at prediabetes, and under 10% reversible once diagnosed.
We treat everyone like the average person.
Identical meals produce dramatically different responses in different people. Population averages cannot guide an individual life.
Same diagnosis, different biology
Two people can carry the same diagnosis and respond to completely different treatments, because the diagnosis describes a symptom, not a cause. Modern psychiatry has no routine way to measure the biology underneath a label, so care runs on trial and error instead of mechanism.
~1 in 3
patients respond to their first depression treatment, despite more than $22B in NIH depression research.
2–6 months
is how long each medication trial takes before anyone knows whether it worked.
21–33%
of patients need two to four trials before finding one that helps.
The fix, in one line
Biotyping resolves one label into the distinct, mechanism-linked patterns hiding inside it, so treatment can be matched to the biology in front of the clinician, not the average patient in the label.
These mechanisms are not neatly boxed by diagnosis either. The same underlying biological factors recur across labels, which is why a genuinely useful map has to work across diagnoses, not one disorder at a time. The reason this gap persists is not neglect: understanding an individual’s disease biology is not something a company can patent, so the step that would fix diagnosis is the step the market has the least incentive to build. That is the market failure Cytognosis exists to close as a nonprofit.
How it works
Four tiers, working as one loop.
Every GPS needs a map, satellites, a coordinate, and a navigator. Cytognosis builds the same four layers for your biology.
Cytoverse turns a diagnosis into a precise position on a map of human biology.
We built a coordinate system for human biology. Instead of sorting people into diagnostic boxes, it gives every person a precise position in a continuous landscape of health, the way latitude and longitude locate a place on Earth.
How this changes care today
- Today: clinicians treat people who share a diagnosis as if they share a cause. Cytognosis: Cytoverse gives each person their own Universal Health Coordinates, precise enough to catch subtle shifts and track them over time.
- Today: diagnosis relies on symptom clusters, with little insight into the biology driving them. Cytognosis: Cytoverse combines detailed health profiles with the genetic and cellular drivers behind them.
Cytoverse · coordinate atlas
Cytograph watches dozens of molecular signals at once, continuously.
A continuous sensor array, from validated wearables and consumer brain sensors to programmable molecular patches. Our sensors reset in under a minute, so they track biomarkers going down as well as up. Most sensors only see them go up.
How Cytograph decides where to look
- Start wide. A broad, low-detail scan establishes a person’s general position on the map.
- Zoom in on what matters. That scan, combined with personal information like genetics, points to the signals worth a closer look.
- Go all the way down, only there. Cytograph zooms to single-cell detail only at those spots, not everywhere.
- Keep adjusting. As health shifts over months and years, Cytograph follows the story wherever it moves.
Cytogram is your live position on the map, updated in real time.
Your live position on the map. Not a label you receive once, but a location with a direction of travel. You see the drift before any threshold is crossed, and you see whether a change is working within days.
Why this tier is new
Earlier versions of this platform had a map, satellites, and a navigator, but never named the thing being computed: your actual position. Cytogram fills that gap, and it is the value that changes minute to minute.
Visualization in progress. A dedicated Cytogram coordinate figure will be added in a follow-up design pass.
Cytorama turns those signals into a specific, personal next step.
On-device AI that turns your coordinate into a route. It finds the root cause rather than the symptom, spots tipping points early, and recommends the specific change that moves you back toward your own baseline. Your raw data never leaves your device.
How this changes care today
- Today: finding the right treatment often means trial and error, based on population averages. Cytognosis: Cytorama matches interventions to each person’s specific biotype.
- Today: clinical decisions mostly wait until a problem crosses a diagnostic threshold. Cytognosis: Cytorama looks for change while it is still reversible, often years before a condition becomes permanent.
map sense locate navigate repeat
Why one scale is never enough
Biology hides at every scale. We look at all of them.
A genetic risk factor, a disrupted brain pathway, and a mood symptom can all trace back to the same underlying cause, but each one is invisible from a different angle. Cytognosis is built to bridge scales that are normally studied in isolation.
Biotypes across scales
Cellular data is the scarce scale: it exists almost only in postmortem brain tissue, so today it is only generated for a handful of disorders. Genomic, connectomic, and phenotypic data are far more available. Rather than wait for cellular data that mostly does not exist yet, Cytognosis estimates the cellular picture from the other scales, and checks that estimate against real tissue as it becomes available. No single scale would let us do that; only having all of them lets one scale fill in for another. This is an illustrative schematic, not real patient data.
Breadth, honestly shown
Fourteen disorders, curated for their biology, not just their name.
Four circuits recur across several of these disorders. Six of the fourteen have no brain region named in our current research, and we show that honestly rather than filling in a guess.
Hover or select a circuit, or click a disorder, to see its associated symptom domain.
See the full multiscale map on Psychoverse .
Cytograph in practice
Watching biology move, years before behavior shows it.
A worked, illustrative example using depression, showing Cytograph’s adaptive sensing in action. Monitoring starts wherever it is easiest to reach, an accessible marker like blood BDNF. As that marker moves in a positive direction, attention does not stop there: it adaptively shifts deeper, first to the brain circuit linked to depression, then, over a longer horizon, to the genomic pattern underneath both.
A diagnosis is a threshold event: before it, the system reports nothing unusual; after it, everything changes at once. A trajectory model tracks the same signal as motion instead, revealing a prevention window that a single snapshot cannot show.
Cytograph · continuous monitoring, years early
Cytograph · worked example, adaptive panel
First deployment
Psychoverse is the platform, built for the brain.
We start with the brain. Neuropsychiatric conditions are where the human cost is highest, where the diagnostic categories fail most visibly, where noninvasive sensing is most mature, and where our team has the deepest track record. Our group built the first single-cell Alzheimer’s atlas (Nature, 2019) and the largest schizophrenia atlas (Science, 2024). Once proven in the brain, the same architecture extends to immune, metabolic, and oncologic disease.
Psychoverse is what Cytoverse becomes when it is pointed at neuropsychiatry: the same map, satellites, coordinate, and navigator idea, built out across genomic, cellular, connectomic, and phenotypic data for mental health. Psychoverse works across four kinds of data, genetic risk, brain and cell biology, brain-circuit activity, and everyday symptoms and behavior, fusing whichever are available for a person and estimating the rest.
Genomic
Inherited risk factors shared across labels.
Cellular
Brain and cell biology, where tissue allows.
Connectomic
Brain-circuit activity and connectivity.
Phenotypic
Everyday symptoms and behavior over time.
Psychoverse
Open, non-diagnostic research infrastructure for studying continuous phenotype dimensions across genomic, circuit, physiological, and behavioral signals.
- Longitudinal context over static labels
- Mechanism-linked, auditable research
- Clear uncertainty and no clinical claims
Yar
Yar, the companion arm of navigation, began inside the Foundation and is forming its own public benefit corporation. It is a private AI companion for neurodivergent minds.
- Private by default and consent-led
- Adaptive, inspectable support
- Roadmap shaped by community votes
How people interact with the platform. Psychoverse’s research map and Yar’s daily companion share the same interpretable coordinates underneath, so the interface always shows why a signal matters, not just that it moved. For the deeper, dedicated overview of this first application area, see neuropsychiatry at Cytognosis .
Open by default
The science belongs to everyone.
Every model, training pipeline, benchmark, and toolkit ships as a public good. Code and weights under Apache 2.0, documentation and non-personal data under CC BY 4.0. We are a 501(c)(3) nonprofit, and the science belongs to everyone.