Quickstart

A compact project layout, reviewed starter parameter file, and recommended command sequence for first-pass evaluation.

Recommended starting point. Begin with a single TOML parameter file in the same working directory as your input structure. Use a short equilibration sequence, confirm the output products, and then scale the production segment after the workflow is behaving as expected.

1. Create a working directory

Structure file + one TOML parameter file

project/
  structure.pdb
  run.toml

2. Start from a documented parameter file

Recommended starter template

[system]
name = "my_system"
pdb_source = "structure.pdb"
output_dir = "outputs/my_system"

[dynamics]
stages = ["nvt", "npt", "production"]

[dynamics.nvt]
steps = 5000
time_step = 0.02
temperature = 300.0
ensemble = "NVT"

[dynamics.npt]
steps = 5000
time_step = 0.02
temperature = 300.0
ensemble = "NPT"

[dynamics.production]
steps = 100000
time_step = 0.02
temperature = 300.0
eval_stride = 50

3. Execute the run

Full workflow from preparation to analysis

neurocgmd run run.toml

This command prepares the system, runs the configured stages, performs CG-to-AA reconstruction, and writes analysis products into the output directory.

4. Inspect outputs

Primary files generated by a standard run

outputs/my_system/
  prepared_bundle.json          # Prepared topology and run context
  energies.csv                  # Energy time series
  traj.jsonl                    # Serialized CG trajectory
  cg_trajectory.pdb             # CG trajectory product
  aa_backmapped_trajectory.pdb  # Back-mapped atomistic trajectory product
  run_summary.json              # Run metadata and summary
  plots/                        # Analysis figures

5. Re-run analysis if needed

Post-process existing trajectory products

neurocgmd analyze run.toml

Use this mode when plots or downstream analysis need to be regenerated without repeating the full simulation workflow.

Recommended review sequence. Keep the parameter file under version control, begin with conservative stage lengths, and confirm that run_summary.json, trajectory products, and analysis figures are being written as expected before extending the production stage.

Parameter-file checklist

Field	What to confirm before launch
`pdb_source`	Points to the intended structure file and matches the working directory.
`output_dir`	Resolves to a clean output location for the current run.
`stages`	Lists the intended order of preparation and production stages.
`steps`	Uses conservative values for the first validation run.
`eval_stride`	Is set deliberately because it controls refinement frequency and runtime cost.

← Return to Home

Open Quickstart Request Evaluation