Tutorial 8 — Continuous Synthesis on a coarse-grained E. coli ribosome (`cosmo-csp`)¶

Goal: synthesize a nascent chain residue by residue on an explicit coarse-grained ribosome, with the elongation cycle split into O’Brien’s three kinetic sub-stages. This is the explicit-ribosome half of the Continuous Synthesis Protocol ported to COSMO’s IDP force field (cosmo.csp, mirroring topo.csp). Its sibling — the same protocol through an analytic cylindrical tunnel — is Tutorial 7.

Unlike topo’s structure-based csp (native contacts, STRIDE, rigid AllBonds), COSMO’s port is sequence-based: no STRIDE, no domain.yaml, no native-contact map — the nascent chain is a transferable HPS/mpipi IDP that grows N→C at the peptidyl-transferase center (PTC). Everything is standalone COSMO: the ribosome is a truncated CG bead model, no external CHARMM files.

This is the tutorial-scaled version of a production run. It uses the same E. coli ribosome, mRNA and O’Brien codon kinetics as the production configuration in sandbox/Ecoli/ (csp_val.ini), but with a short L range, clamped max_steps_per_stage, and CPU so it builds, times, runs and writes outputs end to end in seconds. It is a mechanics demo, not a physical validation — see the production values below.

Prerequisites: the coarse-grained workflow of Tutorial 1 and the force fields of Tutorial 2. The analytic-tunnel sibling is Tutorial 7.

Nascent chain: α-synuclein (asyn.pdb, 140 residues, from Tutorial 1).

The ribosome¶

The ribosome is the E. coli 50S subunit, PDB 4V9D (with the 5JTE A-site tRNA grafted in), coarse-grained to the O’Brien 3/4-bead P/R/BR rRNA representation (4576 beads) and truncated to the region around the exit tunnel: 4v9d_50S_PtR_5jte_AtR_model_cg_trunc.pdb. It enters as rigid (mass-0) scenery: fixed beads that supply the ribosome↔nascent 12-10-6 excluded volume + Yukawa electrostatics wall and the A-/P-site anchors. This is the exact same structure used by the production sandbox/Ecoli/ run.

Each elongation cycle is split into O’Brien’s three kinetic sub-stages — peptidyl-transfer → translocation → tRNA-binding — with the new residue’s C-terminus restraint switching A→P across them, so a residue is delivered to the A site and ratcheted into the P site. The PTC geometry is always optimized: the new residue is seeded one peptide bond (0.380 nm) from the C-terminus, EV-clear.

Files¶

File	Role
`asyn.pdb`	Full-length nascent protein (only the first `L` residues exist at each length).
`4v9d_50S_PtR_5jte_AtR_model_cg_trunc.pdb`	E. coli 4V9D 50S ribosome, CG P/R/BR rep (4576 beads) — rigid scenery + A-/P-anchors.
`mrna.txt`	Demo mRNA for α-synuclein (140 sense codons) — drives the per-codon kinetics.
`trans_times.txt`	E. coli 310 K codon → mean translation-time table (seconds), keyed by codon.
`csp.ini`	3-stage explicit-ribosome config (`model = hps_urry`, real E. coli codon kinetics, tunnel wall on).

Run it¶

cd tutorials/08_csp_cg_ribosome

python -m cosmo.csp -f csp.ini      # -> synth_out_csp/L_<L>/stage_<1,2,3>/
#   installed console script:  cosmo-csp -f csp.ini

The config defaults to device = CPU so it runs anywhere. Kinetics use the real E. coli codon table (trans_times.txt) driven by the transcript (mrna.txt): each residue’s mean time is looked up by codon, split into the 3 sub-stages, and mapped to steps via scale_factor. max_steps_per_stage clamps each sub-stage’s (otherwise ~10⁶-step) dwell down to a traceable test size.

Stitch a movie¶

python -m cosmo.csp.movie -o synth_out_csp --ribosome 4v9d_50S_PtR_5jte_AtR_model_cg_trunc.pdb
#   installed console script:  cosmo-csp-movie -o synth_out_csp --ribosome 4v9d_50S_PtR_5jte_AtR_model_cg_trunc.pdb

--ribosome overlays the static ribosome so you see the chain growing inside it. Writes a fixed-width VMD-playable movie (movie.dcd / movie.psf / movie.tcl).

What it produces¶

synth_out_csp/L_001/ … L_008/, each with stage_1/2/3/ sub-runs — the A→P restraint switch across the three O’Brien sub-stages.
Nascent-only trajectories: each stage_*/traj.psf/.dcd contains only the L nascent beads — the 4576-bead ribosome is excluded from the trajectory and re-attached as static scenery by the movie stitcher’s --ribosome overlay.
synth_out_csp/dwell_times.dat — per-residue dwell-time table.

Generated synth_out_*/ directories are git-ignored (bulky trajectories/checkpoints). Delete them and re-run to reproduce.

Going to production¶

The full production run this tutorial is scaled down from lives in sandbox/Ecoli/ (csp_val.ini). To reproduce it, start from this csp.ini and:

Full length — remove L_max (synthesize all 140 residues).
Un-clamp the dwell — remove max_steps_per_stage / min_steps_per_stage so each stage’s step count comes from the real dwell-time calculation.
GPU — device = GPU, ppn = 4.
Post-synthesis — ejection_steps = 10_000_000 (let the finished chain diffuse out of the exit tunnel).

The same nascent chain + kinetics are run against different organisms’ ribosomes in sandbox/ (Ecoli / Human / Yeast / Ncrassa) — only the ribosome PDB changes — to compare co-translational behaviour across ribosome structures.

Where this sits in the series¶

The two co-translational tutorials split cosmo.csp by confinement geometry: Tutorial 7 is the analytic cylindrical tunnel (no ribosome beads; fast, never jams), and this one is the explicit coarse-grained ribosome with O’Brien’s 3-stage per-codon kinetics.

Tutorial 8 — Continuous Synthesis on a coarse-grained E. coli ribosome (cosmo-csp)¶