Molecular machines don’t move from a dropdown. Any saved selection becomes a component; behaviours stack like clips; relationships wire components into machinery — all of it editable, all of it stored beside the data.
No-code · Unconstrained
Build motion from anything
Turn any saved molecular selection into a component, then stack general behaviours and relationships. Every layer remains independent and editable.
New component
Start anywhere. Protein domains, ligands, DNA strands, waters, or a custom residue query can all be components.
Behaviours repeat in phased cycles; components generate time-offset copies of themselves; and any component can attach to — or detach from — another moving component. Wire a ratchet to a rotor to a drifting cargo and you have a machine.
The example constructions ship as conveniences, not fixtures — each compiles to the same editable component graph you would build by hand, and remains optional. Take one apart to learn; rebuild it your way.
Axes and pivots derive from the selection itself — a helix axis, a hinge — or float free where you place them. The geometry of the data guides the motion; it never limits it.
All motion lives in the client-side project document. Deposited atom coordinates remain unchanged — and provenance labels keep demonstration clearly separated from evidence.