The ribosome structure (get one, or build your own)

The explicit-ribosome runner (Synthesis in coarse-grained ribosome model) needs a rigid, oriented, truncated, coarse-grained large subunit as its ribosome input — the ribosome_trunc.pdb you see in the tutorials. TOPO ships ready-made ones for the most-studied organisms, and includes the scripted pipeline that builds them, so you can also prepare your own from any PDB deposition. (The cylinder runner needs none of this.)

Every structure is oriented in the TOPO tunnel frame: PTC at the origin, exit tunnel on +x, tRNA tails on +y — so the tunnel axis is the X-axis and the radial distance to it is sqrt(y²+z²).

Ready-made reference ribosomes

They live in the repo at assets/csp/prepare_ribosome/structures/<organism>/, two files each: *_model_cg.pdb (the full CG large subunit — a re-truncation master) and *_model_cg_trunc.pdb (the default exit-tunnel crop, ready to run).

Organism	Large subunit / PDB	PTC landmark	Exit landmark (uL24)	Trunc beads
E. coli	50S / 4V9D (+5JTE A-tRNA)	23S:2602:N6	L24:51:N	4 575
S. cerevisiae	60S / 6Q8Y	25S:2971:N6	L26:91:N	5 199
N. crassa	60S / 7R81	26S:2931:N6	L26:91:N	5 578
H. sapiens	60S / 8G61	28S:4548:N6	L26:93:N (RPL26)	5 795

Per-organism provenance (chain→segID maps, landmark confirmation, grafts) is in each structures/<organism>/PROVENANCE.md; the full catalog is MANIFEST.md.

Note

Caveats to cite when you use these. The yeast A-site tRNA is a cross-species graft from the human 8G61 decoding complex (6Q8Y has only P/E tRNAs); its acceptor is a sane A-site anchor but not a native yeast tRNA. The eukaryote exit landmark (uL24 = RPL26 = segID L26, not eL24) is homology-assigned (±2 residues on the vestibule loop; orientation is insensitive to this). N. crassa uses its native A/P tRNAs (its A-site is not empty); its bound nascent peptide is dropped.

Use one in a run

The ribosome key of a csp.ini is a plain path — point it at a truncated structure:

pdb_file  = my_protein.pdb
ribosome  = assets/csp/prepare_ribosome/structures/human/8g61_60S_model_cg_trunc.pdb
domain_def = my_protein.yaml

Re-truncate with your own criteria

The default crop keeps the exit-tunnel shell (r_cyl 30 Å, x_lo −8, x_exit 58). To keep more/less of the wall, a longer exit cap, or the whole tRNAs, run the truncator on the shipped full CG (already oriented, so the X-axis is the tunnel):

cd assets/csp/prepare_ribosome
python truncate_ribosome.py \
    -i structures/human/8g61_60S_model_cg.pdb -o my_human_trunc.pdb \
    --r-cyl 35 --x-lo -8 --x-exit 70 --keep-segids PtR,AtR

Build a new organism

The pipeline is five stages — acquire → gen → orient → coarse-grain → truncate — driven by run.py:

cd assets/csp/prepare_ribosome
bash fetch_cifs.sh                     # download raw cifs (needs internet; gemmi)
python run.py -c configs/<organism>.ini

Adding a new organism means writing its config (landmark residues + chain→segID map) and verifying the PTC and exit landmarks by structural homology — do not guess, since rRNA/protein numbering differs between organisms. The bundle’s RUNBOOK.md is the methodology + “how to add an organism” guide, and helpers/scan_landmarks.py / verify_landmarks.py confirm the landmarks against the validated E. coli frame. See assets/csp/prepare_ribosome/README.md for the full workflow.