Biology
Our biology platform, BIOMOL, applies physics intelligence to records that are genuinely dynamical: molecular-dynamics trajectories and time-course proteomics. We chose biology because the data is rich, the ground truth is published, and the field’s established methods are statistical. The platform rests on one principle, which we call structure control: a molecule’s structure determines what it can do, and the structure can be read by physics before the behaviour is observed. The first test is deliberately modest: can a physics reading, with no biological knowledge and no fitted parameters, recover findings that biology has already established? Recovery is the validation. Only then does the interesting question follow — what physics sees that established methods do not.
Three tracks are running. FlexSense reads a molecular-dynamics trajectory and reports whether the molecule’s dynamic regime has changed, with an inconclusive state where the record does not decide. StabilityLens asks whether the platform can rank the stability of variants within a protein family; its first pilots returned negative results under their pre-registered criteria, and those results stand on the record. A third study applied the reading to a published time-course of cellular reprogramming: the strongest transition it found falls in the interval the literature identifies as the intermediate phase, an interval conventional distance metrics do not single out, while a pre-registered enrichment analysis on the same data returned negative. Both outcomes are reported.
The structural-triage layer at protein.neoamorfic.ai is the first, basic application of structure control and a building block of the platform: it reads a protein’s structure and reports what that structure admits. The evidence for each study is published with the study. We run no wet laboratory and make no clinical claim, and every study is contract-first: scope, settings and acceptance criteria are frozen and hashed before a calculation is run. We are open to working with structural-biology and molecular-dynamics groups who want an independent physics reading of their own trajectories.