Configuration Files¶
Configuration files are one of the main ways AToM-OpenMM workflows are controlled. The newer tutorials use YAML files because they are readable, easy to copy between related calculations, and can be combined with system-specific settings generated by the setup scripts.
AToM-OpenMM can still read .cntl, .yaml/.yml, and .json files. Some older tutorials use *_asyncre.cntl directly; they remain useful examples, but new workflow documentation should focus on the YAML-based pattern.
Newer YAML Workflow¶
The newer ABFE and RBFE tutorials use this structure:
scripts/
setup-settings.sh
defaults.yaml
run-atm.py
run_template.sh
complexes/
system-name/
system-name.yaml
run.sh
The roles are:
| File | Role |
|---|---|
setup-settings.sh |
Lists the systems or perturbations to prepare, such as receptors, ligands, peptide pairs, mutation residues, or alignment inputs. |
defaults.yaml |
Stores shared simulation options used across all systems in the tutorial. |
run-atm.py |
Combines the shared defaults with system-specific information and generated atom selections. |
system-name.yaml |
The generated per-system options file written into each calculation directory. |
run.sh |
Launches preparation, production, and analysis for one system or perturbation. |
For most edits, start with scripts/defaults.yaml. It controls the common settings shared across the calculation set. The YAML files include comments, so the file itself is often the most compact reference.
Keyword Breakdown¶
The tables below explain the main keywords used by the newer defaults.yaml files and the generated per-system YAML files. The YAML comments remain the quickest reminder while editing; this section adds context and notes about which workflows use each setting.
Alchemical schedule
| Keyword | Meaning |
|---|---|
TEMPERATURES |
Simulation temperature list. Most examples use one temperature. |
LAMBDAS |
Labels the alchemical states and determines the number of replica states. |
DIRECTION |
Direction of each alchemical state. Positive and negative directions define the two legs of the ATM cycle. |
INTERMEDIATE |
Marks intermediate states, commonly around the midpoint of the alchemical path. |
LAMBDA1 |
First softplus lambda parameter for each state. |
LAMBDA2 |
Second softplus lambda parameter for each state. |
Soft-core and bias parameters
| Keyword | Meaning |
|---|---|
ALPHA |
Softplus sharpness parameter. |
U0 |
Softplus energy offset parameter, in kcal/mol. |
W0COEFF |
Additive bias coefficient for each state. |
UMAX |
Maximum energy parameter for the soft-core binding energy, in kcal/mol. |
ACORE |
Dimensionless soft-core parameter. |
UBCORE |
Soft-core energy parameter, in kcal/mol. |
Runtime and sampling
| Keyword | Meaning |
|---|---|
WALL_TIME |
Requested production runtime in minutes. Keep this consistent with scheduler limits. |
MAX_SAMPLES |
Target maximum number of perturbation-energy samples per replica. Tutorial values are intentionally small. |
CYCLE_TIME |
Time interval between replica-exchange scheduling cycles, in seconds. |
CHECKPOINT_TIME |
Time interval between checkpoint writes, in seconds. |
SUBJOBS_BUFFER_SIZE |
Size of the fast execution queue, expressed as a fractional number relative to the number of GPU devices. |
PRODUCTION_STEPS |
Number of MD steps each replica runs per production segment. |
Output frequency
| Keyword | Meaning |
|---|---|
PRNT_FREQUENCY |
Interval, in MD steps, for printing replica data. |
TRJ_FREQUENCY |
Interval, in MD steps, for trajectory frame output. |
VERBOSE |
Enables more detailed output when supported by the workflow. |
Dynamics and force fields
| Keyword | Meaning |
|---|---|
FRICTION_COEFF |
Langevin friction coefficient. |
HMASS |
Hydrogen mass used for hydrogen mass repartitioning. |
TIME_STEP |
MD time step, in ps. |
LIGAND_FORCE_FIELD |
Force field used for the physical ligand. |
Binding-site and positional restraints
| Keyword | Meaning |
|---|---|
RCPT_CHAIN_NAMES |
Receptor chain IDs used to build the receptor frame and identify receptor atoms. |
CM_KF |
Force constant for the flat-bottom binding-site center-of-mass restraint, in kcal/mol/Angstrom squared. |
CM_TOL |
Radius/tolerance of the flat-bottom binding-site restraint, in Angstroms. |
POSRE_FORCE_CONSTANT |
Force constant for positional restraints, in kcal/mol/Angstrom squared. |
POSRE_TOLERANCE |
Radius/tolerance for positional restraints, in Angstroms. |
Workflow-specific keywords
ABFE with a ghost ligand:
| Keyword | Meaning |
|---|---|
GHOST_MASS |
Mass assigned to the one-atom noninteracting ghost ligand. |
Small-molecule RBFE:
| Keyword | Meaning |
|---|---|
LIGAND_FORCE_FIELD |
Force field used for ligands and cofactors, such as OpenFF or GAFF. |
ALIGN_KF_SEP |
Separation component of the ligand alignment restraint. A value of zero is used in variable-displacement examples. |
ALIGN_K_THETA |
Orientation component of the ligand alignment restraint. |
ALIGN_K_PSI |
Roll component of the ligand alignment restraint. |
ALIGN_LIGAND1_REF_ATOMS |
Generated reference atom indexes for ligand 1 alignment. |
ALIGN_LIGAND2_REF_ATOMS |
Generated reference atom indexes for ligand 2 alignment. |
Protein-peptide RBFE:
| Keyword | Meaning |
|---|---|
MUTATION_RESID |
Generated mutation residue identifier used to define common and variable peptide atoms. |
PP_BACKBONE |
Optional peptide backbone atom names used by the protein-peptide helper. |
ALIGN_KF_SEP, ALIGN_K_THETA, ALIGN_K_PSI |
Alignment restraint settings used by the RBFE production workflow. |
System Specific keywords
These settings regulate in detail the nature of the alchemical transformations. They are usually generated automatically by the workflow scripts.
| Keyword | Meaning |
|---|---|
BASENAME |
Job name used to derive input and output filenames. |
WORKDIR |
Calculation working directory. |
DISPLACEMENT |
Vector that places the second ligand, peptide, or ghost in the displaced solvent position. |
LIGOFFSET |
Offset used internally for ligand placement and ATM setup. |
LIGAND1_ATOMS, LIGAND2_ATOMS |
Atom indexes for the two alchemical partners. In ghost-ligand ABFE, ligand 2 is the ghost. |
LIGAND1_VAR_ATOMS, LIGAND2_VAR_ATOMS |
Variable atom indexes affected by the alchemical transformation. |
LIGAND1_ATTACH_ATOM, LIGAND2_ATTACH_ATOM |
Attachment atoms used for displacement and visualization helpers. |
LIGAND1_CM_ATOMS, LIGAND2_CM_ATOMS |
Atom indexes used to define ligand centers of mass. |
RCPT_CM_ATOMS |
Receptor atoms defining the binding-site center. |
RCPT_FRAME_ATOMS_O, RCPT_FRAME_ATOMS_Z, RCPT_FRAME_ATOMS_Y |
Receptor atom groups defining the internal receptor frame. |
POS_RESTRAINED_ATOMS |
Atom indexes subject to positional restraints. |
EXCLUSION_POT_MOL1_INDEXES, EXCLUSION_POT_MOL2_INDEXES |
Atom indexes used by the short-range exclusion potential. |
Tutorial defaults are usually set for short demonstration runs. For quantitative work, increase MAX_SAMPLES, WALL_TIME, and scheduler time limits consistently.
Generated Per-System YAML¶
The helper workflows write a merged YAML file into each calculation directory. This file records the final options used by rbfe_structprep, rbfe_production, and related analysis calls.
When debugging, compare:
scripts/defaults.yaml: shared settings you intended to use.scripts/setup-settings.sh: systems, pairs, and workflow-specific setup inputs.complexes/<system>/<system>.yaml: final generated options for one calculation.- The log file: echoed runtime options and progress messages.
In general, edit defaults.yaml for shared simulation behavior and edit setup files for system lists or pair definitions. Treat generated per-system YAML files as records of a specific prepared calculation unless you are deliberately debugging or restarting that calculation.