Psi4 Support Mode

easyPARM v4.10 now supports the Psi4 Open-Source Quantum Chemistry package. All you need is your .xyz structure file — easyPARM will handle the rest: geometry optimization, frequency analysis, RESP charge calculation, and parameter generation for your complex.

Tutorial

1. Installation

You need Conda, which is the easiest and most reliable way to install Psi4.
You don’t need the full AMBER package — only AmberTools, which is available via Conda.

# 1. Download miniconda
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh

# 2. Install miniconda
chmod +x Miniconda3-latest-Linux-x86_64.sh
./Miniconda3-latest-Linux-x86_64.sh

# 3. Initialize conda
source ~/miniconda3/etc/profile.d/conda.sh

# 4. Create Psi4 environment
conda create -n psi4env psi4 -c psi4

# 5. Activate environment
conda activate psi4env

# 6. Install AmberTools (only AmberTools is needed)
conda install conda-forge::ambertools

# 7. Install RESP method (Alenaizan’s implementation)
conda install resp -c psi4
conda install pytest

# 8. Install required packages
conda install periodictable biopython 

After installing all required packages, you are ready to run easyPARM with Psi4 support.

2. Running easyPARM

It is recommended to run easyPARM in non-interactive mode to fully use Psi4 features.

In order to use the non-interactive mode, check the dedicated section and prepare an input file for easyPARM. This input must include variables related to Psi4, such as:

1. CHARGE_METHOD and QM_OUTPUT → set = 6 for Psi4.
2. Additional Psi4-specific settings:

# PSI4 SETTINGS (QM_OUTPUT = 6)
CPU = 24                       # Number of CPU cores
MEMORY = 50GB                  # Memory allocation (e.g., 5GB, 500MB, 1000MB)
METHOD = B3LYP                 # Functional (default: B3LYP)
BASIS_SET = 6-31G*, DEF2-SVP   # Basis sets (default: 6-31G* for non-metals, LANL2DZ for metals)

A key control parameter is:

# PSI4 = ON  -> run a new calculation
# PSI4 = OFF -> skip (use existing outputs: resp_charges.dat, hessian.txt, optimized.xyz)

For example, if you have already run Psi4 once, you can reuse resp_charges.dat, hessian.txt, and optimized.xyz with PSI4 = OFF to avoid repeating the calculation.

3. Execution Methods

Method 1 (Local)

Run directly:

/path/to/easyPARM.sh parameter.input

⚠️ Not recommended for large systems (may take significant time).

Method 2 (Cluster/PBS)

Submit via cluster scheduler (PBS example):

#!/bin/bash
#PBS -S /bin/bash
#PBS -m be
#PBS -u your_username
#PBS -N _easyPARM
#PBS -l nodes=qcexnod56:ppn=24

set -x
export HomeDir=$PBS_O_WORKDIR

# Scratch directory
export WORKDIR=/scr/your_username/${PBS_JOBID}_$JOB
export PSI_SCRATCH=$WORKDIR
mkdir -p $WORKDIR

# Activate Psi4 environment
source ~/miniconda3/etc/profile.d/conda.sh
conda activate psi4env

# Copy inputs
cp $HomeDir/your_structure.xyz $WORKDIR
cp $HomeDir/your_input $WORKDIR
cd $WORKDIR

# Run job
/path/to/easyPARM.sh your_input > $HomeDir/output.dat

# Copy results back
cp $JOB.log $HomeDir/
cp COMPLEX.* $HomeDir/
cp *.dat $HomeDir/
cp hessian.txt $HomeDir/

# Optional cleanup
#rm -rf $WORKDIR
exit

Modify this script according to your cluster configuration.

4. Output Files

After completion, you will obtain both AMBER parameter files and Psi4 results:

1. hessian.txt – Hessian matrix (2nd derivatives of energy)
2. resp_charges.dat – RESP charges
3. optimized.xyz – Optimized geometry
4. calculation_summary.txt – Calculation summary (functional, basis set, CPUs, etc.)

These files can be reused with PSI4 = OFF to avoid recalculating.

5. Advanced Modifications

If you are familiar with Psi4, you can modify calculation settings in PSI4.py:

The following options provide a balanced approach between computational cost and accuracy.

# Psi4 calculation options
    psi4.set_options({
        'reference': 'UHF' if multiplicity > 1 else 'RHF',
        'scf_type': 'DF',
        'maxiter': 300,
        'guess': 'SAD',
        'e_convergence': 1e-8,
        'd_convergence': 1e-7,
        'g_convergence': 'GAU',
        'geom_maxiter': 300,
        'dft_nuclear_scheme': 'TREUTLER'
    })

Defaults use DF method and GAU convergence for balance cost and efficiency. You may adjust these parameters depending on your needs. For example, you could modify the auxiliary basis or use GAU_TIGHT or increase the threshold for energy and density convergence.

Please report any Psi4-related bugs — this is the first release with Psi4 support.