config.yaml

#######
# RUN #
#######
# Initialization parameters
input: akg_nhie.csv                    # Path to protein csv, single PDB, or list of PDBs [PDB1, PDB2, PDB3]
output_dir: dataset/v1                 # Path to save the generated clusters
modeller: true                         # Model missing loops
protoss: true                          # Assign protonation states
coordination: true                     # Generate clusters
skip: all                              # Skip 'modeller', 'protoss', or 'all' jobs if they already exist
optimize_select_residues: 1            # Modeller to model no (0), missing (1), or all (2) residues
convert_to_oxo: true                   # Used to convert A-KG NHIEs to Succinate-Oxo NHIEs, usually False
max_clash_refinement_iter: 5           # The max number of times to iterate over the modeller protoss loop 
# Cluster model parameters
max_atom_count: 750            
center_residues: [FE, FE2]             # Residue name to use as the cluster center
additional_ligands: []                 # Residues beyond first sphere to include
number_of_spheres: 3                   # Coordination spheres to generate
radius_of_first_sphere: 4.0            # Distance based cut off to determine first sphere
include_ligands: 2                     # Only ligands and waters in first sphere (0), Only non-water ligands (1), Everything (2)
capping_method: 1                      # Cap residues with None (0), hydrogens (1), ACE/NME (2)
smoothing_method: 2                    # Smoothing method Box plot (0), DBSCAN (1), [Dummy Atom] (2), None (3)
# Output parameters                               
compute_charges: true                  # Calculate charges using Protoss output
count_residues: true                   # Count the number of residues
write_xyz: true                        # Write out xyz files

##########
# SUBMIT #
##########
# Create QM job files
optimization: false                    # Optimization (true) or single point (false)
method: ub3lyp                         # Functional to use
basis: lacvps_ecp                      # Basis set to use
guess: generate                        # Method for initial guess
gpus: 1                                # How many GPUs to use
memory: 4G                             # How much memory to use
dielectric : 10                        # The dieletric to use if using implicit solvent  
charge_embedding: true                 # Prepare QM with charge embedding
scheduler: slurm                       # What scheduler do you use (e.g., slurm, sge)
pcm_radii_file: /path/to/pcm_radii     # The path to the pcm_radii file if using terachem
# Submit QM jobs
create_jobs: false                     # Create QM job input files
submit_jobs: false                     # Submit the created jobs
job_count: 10                          # The number of jobs to maintain on the scheduler

###########
# ANALYZE #
###########
job_checkup: false                     # Check how status of qm jobs
delete_queued: false                   # Clean up jobs by deleting .submit_records for unfinished jobs
atmrad_path: /data1/groups/HJKgroup/src/multiwfn/Multiwfn_3.7_bin_Linux_noGUI/examples/atmrad
charge_scheme: Hirshfeld,CM5           # Choose from the following: Hirshfeld, Voronoi, Mulliken, ADCH, Hirshfeld-I, CM5
calc_charge_schemes: false             # Calculate charge schemes with Multiwfn
calc_dipole_com: true                  # Calculate dipole with Multiwfn using the center of mass as the reference
calc_dipole_coc: false                 # Calculate dipole with Multiwfn using the center of charge as the reference