Three configurations driven by the actual MEDYAN C++ binary at
/path/to/medyan through MedyanCxxProcess.
All filament beads, motor and linker endpoints, and
membrane mesh vertices shown below come straight
from MEDYAN’s snapshot.traj and traj.h5
outputs — this is the upstream simulator, not a Python
reimplementation.
Actomyosin contractile collapse vs MD copy number (Popov et al. 2016)
Four runs of the same actomyosin network at varying myosin-II minifilament densities (MD diffusing-species copy = 0 / 5 / 20 / 50). With more available motors, more bind to filaments per MEDYAN's low-duty-cycle catch-bond reactions and pull on neighbouring cylinders along the filament axis, contracting the network. The reported network span (bead bounding-box diagonal) is a classical MEDYAN metric for the contractile transition.
Polymerizing actin pushes a deformable lipid vesicle outward
A closed lipid vesicle (icosphere mesh, 1000+ vertices, Helfrich bending + constant tension + volume conservation) wraps a small actin network. Brownian-ratchet polymerization extends filament plus-ends; tip contact with the membrane is mediated by MEDYAN's triangle–bead repulsion force-field. The membrane geometry shown comes straight from snapshots/<i>/membranes/0/{vertexDataFloat64,triangleDataInt64} in MEDYAN's HDF5 trajectory. This is the flagship MEDYAN-vesicle subsystem in action.
A separate PBG process drives MEDYAN's G-actin concentration
The wrapper exposes an actin_copy input port. On each update(state, interval), a sibling PBG process can push a new G-actin (AD) diffusing-species copy number; the wrapper rewrites the chemistry input and re-runs MEDYAN with the new value. Here the schedule is a square wave 800 → 100 → 800, simulating a regulatory pulse. Above the critical actin concentration the network grows; below it, koff dominates and filaments shrink. This kind of cross-process coupling is exactly what bare MEDYAN can't do — the wrapper turns it into a one-line input wiring in any larger Composite.