Curing complex diseases requires combinations.

Drug discovery is organised around finding one target that works on its own. Biology is not built that way.

Biology runs on redundancy and feedback, so blocking one route very often means the cell reroutes through another. That is the core problem in cancer resistance and regrowth, and it applies across complex evolving diseases more widely, in immune disease, in neurological disease, and where the two intersect.

Oncology has already started combining: around one in three FDA solid-tumour approvals between 2011 and 2023 were combination regimens.

We strongly believe we probably have enough drugs, enough identified targets, and honestly probably enough large-scale data. What is not known is which combination, and in which sequence, produces an optimal synergistic effect that is sustained. That is a representation problem and a search problem at the same time, intractable with typical drug screens, and tractable with the right mix of agentic systems and fast dynamical models.

The scale is the reason. A pairwise screen of 100 drugs at 100 doses each runs to roughly 50 million experiments, and a dish still cannot see the immune compartment clearing a clone, the stroma walling a drug out, how much of a dose reaches each metastatic site, or what changes when you reverse the order of two agents.

In our first few months we have found that despite enormous genomic variability, resistance pathways across cancers share a similar set of strategies, and a similar order for using them. That opens the possibility, for the first time, that a relatively small set of the right drugs could move some of the worst prognoses from around 12 months to decades.

Read the full mission statement