Epigenetic reprogramming with transient OSK. The goal is to reset age markers in tissue without destabilising cell identity.
Modelling
Core EPV-0412Depth 62.4 mΔ Methylation −0.148Zero EgressOn-premise
Epivora Labs
Independent AI biotech laboratory
Horsens · DK
The
Longevity
Dividend
Biological age isn’t fate. It’s a reading — laid down layer by layer inside the cell, the way years are laid down in an ice core. Readings can be changed. We work to extend the healthy years, before demography catches up with the health system.
Four research tracks · one instrumented pipeline
01The Demographic Winter
Winter isn’t a forecast.
It’s already booked.
The people who will fill the care system in 2045 have already been born. The number isn’t up for negotiation. The only thing still open is how many of their years are healthy ones.
Health systems are built to repair damage after it happens. That model doesn’t scale once the curve turns. The alternative isn’t more hospitals — it’s moving the intervention back to the cell, before the pathology is established.
This is what we call The Longevity Dividend: every healthy year recovered is both a person kept out of the system and a working life society gets to keep. Health becomes infrastructure, not expenditure.
0
Years is the rough gap between life expectancy and healthy life expectancy. That gap is where we work.
0
The growth in the 80+ population in Denmark by the mid-2040s. Care capacity does not grow to match.
0
Of the disease burden in older adults traces back to a handful of ageing-related mechanisms. One mechanism, many diagnoses.
Implication
Treating ageing as a risk factor — not as background noise — is the only lever that acts on several diagnoses at once.
Figures are indicative and must be verified against Statistics Denmark and WHO HALE before publication.
02The Core · Biological age
We read age
like an ice core.
An ice core tells you which year every layer was laid down. The epigenome does the same: methylation patterns accumulate over time and can be read as an age — often a different age than the one in the passport.
The difference is that ice can’t be relaid. The pattern can. Transient reprogramming aims to reset the accumulation without touching cell identity — maintenance, not transformation.
Stratigraphy · EPV-0412
Biological age
58
Δ 0.0 yrs
Illustrative model — not clinical data
Method
Read before you intervene. Every hypothesis has to be measurable on an epigenetic clock before it becomes a protocol. No protocol without a baseline.
Principle
Transient means time-limited. The cell should remember what it is — and only forget how long it has been that.
03Four Pillars · Research tracks
Four tracks. One instrument.
The tracks share one pipeline, one literature index and one pool of compute. That is what makes a small lab faster than a large one: no cost to move a finding from one track to another.
Personalised mRNA and peptide design aimed at the individual’s inflammation profile rather than an average patient.
DesignMetabolic oncology. The Warburg effect attacked with triple agonists — starve the tumour instead of poisoning the host.
Literature synthesisOncolytic virotherapy. A virus as a targeting agent that also makes the tumour visible to the immune system.
Literature synthesis04The Instrument · Infrastructure
The brain stays
in the building.
We build the lab in the order that makes sense for a small team: the reading first, the hands second. The brain is already running — a cluster that cross-reads global medical research around the clock and keeps every finding on our own disks.
Zero Egress
No patient data leaves the building. No public clouds in the research loop. Data sovereignty isn’t a feature — it’s the precondition for being allowed to work with the data at all.
ComputeHPE DL380 Gen10 clusterBuilt from the ground up. On-premise, our own power, our own access control.
IndexQdrant vector databasesLiterature, preprints and datasets made searchable by meaning rather than keywords.
PipelineAsynchronous AI workflows in n8nRuns in the background, reports anomalies. People read conclusions, not logs.
NextOxford Nanopore · OpentronsThe hands: sequencing and automated liquid handling in our own cloud lab.
05The Ascent · The capital ladder
Four rungs to
an automated lab.
The ladder is built so every rung stands on its own. If only the first one is funded, the lab still gains a capability it didn’t have yesterday. From the bottom up:
Full Build
Redundancy + expanded computeFull capacity: duplicated critical hardware, an expanded cluster and room to run all four tracks in parallel.
2,400,000DKK
Rung 03
Integration + runwaySequencing and robotics linked into one chain — plus the operating runway to keep it going while it validates itself.
1,400,000DKK
Rung 02
Opentrons liquid handlingThe hands. Automated pipetting gives a reproducibility no human can deliver at scale.
750,000DKK
Rung 01
Oxford Nanopore sequencing benchThe reading. Without in-house sequencing everything else is theory — this rung makes the lab wet.
280,000DKK
06Three ways to build this lab
Each track has its own page,
its own ask, and its own answer.
01The Brain
Enterprise compute, memory, storage and networking for on-premise training and inference.
Compute → 02The HandsAutomated liquid handling, sequencing and assay hardware to close the research loop.
Wet Lab → 03The FuelCapital for physical rungs of capability and sustained model inference.
Capital →