What COSMO does

COSMO builds a one-bead-per-residue, sequence-based coarse-grained model of intrinsically disordered proteins (and RNA) and runs Langevin dynamics on OpenMM. Interactions come from the sequence — the hydropathy-scale HPS (Ashbaugh–Hatch) or the mpipi (Wang–Frenkel) force field plus Debye–Hückel electrostatics — not from a folded structure, so COSMO is built for disordered chains: single-chain dimensions, phase separation, and complexes.

That model powers two complementary workflows:

• A. Coarse-grained simulation of IDPs — single chains, multi-chain slabs / LLPS, and protein–RNA complexes.

• B. Protein synthesis — grow the chain residue by residue on the ribosome under codon-resolved kinetics, and watch how a disordered chain extrudes and behaves as it is made.

Part B builds on the Part A force field, so start with A if you are new here.

A. Coarse-grained simulation of IDPs

Start from a sequence (a CA/P PDB) and run HPS / mpipi dynamics: single-chain radius of gyration, temperature/pressure coupling and PBC, slab simulations of liquid–liquid phase separation, and protein–RNA mixtures.

Tutorials

• 1 · Single-chain quickstart

• 2 · Models & force fields

• 3 · PBC, temperature & pressure

• 4 · Restart & outputs

• 5 · Slab / LLPS

• 6 · Protein–RNA complex

Reference

• Simulation control options (md.ini)

B. Protein synthesis

Grow the nascent chain N→C, one residue at a time, timing every residue from its mRNA codon (the O’Brien Continuous Synthesis Protocol, ported to cosmo’s IDP force field). Two runners differ in how the ribosome exit tunnel is represented:

• ``cosmo-cylinder`` — the exit tunnel is an analytic cylindrical bore through an infinite wall (no explicit ribosome beads); fast, never jams, one MD segment per residue.

• ``cosmo-csp`` — the ribosome-based counterpart: grow the chain through an explicit truncated CG ribosome PDB (see The ribosome structure (get one, or build your own)) with the O’Brien 12-10-6 excluded volume, in three codon-timed sub-stages per residue, then eject the completed chain.

Which to use. Reach for the cylinder for fast exploration of how tunnel geometry + codon kinetics shape a disordered chain as it extrudes, or when you have no ribosome structure — it is the simplest starting point. Reach for the explicit ribosome when the tunnel-wall charge, the real tunnel shape (constriction / vestibule), or translocation-coupled forces matter to your question.

Warning

The two runners are comparable only in the *mean* per-residue dwell time, not in confinement chemistry. The cylinder omits the ribosome’s electrostatics and surface excluded volume and has a uniform straight bore. Do not compare extrusion/conformation observables (radius of gyration vs. length, contact formation) across the two runners without accounting for those missing terms.

Runnable proof-of-concept examples (α-synuclein) live in sandbox/validate/ (csp.ini and cylinder.ini).

Tutorials

• 7 · Continuous synthesis — analytic tunnel (cosmo-cylinder)

• 8 · Continuous synthesis — coarse-grained E. coli ribosome (cosmo-csp)

Reference

• The ribosome structure — get or build one

• Synthesis on a coarse-grained ribosome

• Synthesis through an analytic tunnel

• Synthesis control options (csp.ini)

• cosmo.csp API reference

Note

Protein synthesis is the cosmo port of the sibling topo package’s topo.csp. It shares the codon kinetics, the three-stage protocol, the cylinder model, and the O’Brien 12-10-6 ribosome excluded volume, but the nascent chain is a sequence-based IDP (HPS / mpipi) rather than a structure-based Gō model — so there is no STRIDE, native-contact map, or domain.yaml.