Based in San Carlos, Precigenetics was founded by hardware engineers, computational biologists, and cell biologists to make drug discovery truly relevant to human biology. Today, I announce our 2026 plan: extending our measurement into a foundational dataset for cell states.🧵

Precigenetics is building a biophotonics–AI engine to understand cell states across space and time. Our mission is beyond AI, beyond disease, it is a matter of digitizing and understanding biology. An AI Virtual Cell atlas requires a universal representation.

AI does not even have the language to understand - the universal representation - to grasp, the complexities of a living, human cell. Our hardware provides the missing layers.

It is time to go all-in, and create datasets that change how we interact with living biology with a computer. Layers of -omics go missing with our traditional measurement apparatus - we have unlocked metabolomics, lipidomics, and some forms of epigenomics in live cells.

Milestone: Build a foundational dataset for dynamic human cell states: live-cell trajectories, thousands of drug–cell and drug–model interactions, across patient-derived 3D cultures and organ-on-chip systems.

Mapping how drugs reprogram living cells to build a dynamic model of the human cell is the most important work of this lifetime.

Our platform is built on advances in label-free optics, microfluidics, and AI, and records non-destructive, hyperspectral videos of living human cells as they respond to drugs, gene edits, and microenvironmental changes.

From each trajectory, we extract rich biochemical fingerprints aligned with omics and outcomes, learning how cell states evolve and which interventions actually matter.

This data, and the models trained on it, make it easier to design combination therapies, de-risk toxicity, and uncover whole new classes of disease biology - starting with oncology.

We treat every experiment as a movie, not a snapshot: thousands of time points per cell, subcellular spatial resolution, and controlled perfusion of drugs and stimuli.

This lets us see early mitochondrial, lipid, and redox flux; the precursors of efficacy, resistance, or toxicity, long before traditional endpoints.

We are now building virtual cell-state models: foundation models pre-trained to learn how genes, pathways, and environments interact in living cells.

Instead of learning only from static RNA or protein levels, our models learn from full chemical trajectories.

By aligning these optical trajectories with RNA-seq, epigenomics, and clinical context, we build representations that can simulate “what if” interventions and predict how a cell will respond before the experiment is run.

We believe our new milestone will be the most important mission we have ever embarked, and we are ready to take it on today, and for all of 2026.

Please share this with people who may be interested in such an effort. We aim to make biology truly digitised across time, and when it comes to understanding disease, we believe we are truly at the start of a revolution. It’s time to build.