The_EID_charge_correction_method

This repository contains the implementation of the Electrostatic Interaction Decoupling (EID) workflow for charge-changing alchemical free energy calculations.

In the current EID formulation, the strict statement applies to the isolation of the linear term from the quadratic fit, whereas the final electrostatic free energy is obtained by adding a residual offset-removal analytical correction to the linearly separated periodic-boundary-condition (PBC) contribution. The practical workflow therefore contains two parts:

1. extract the linearly separated PBC contribution from the quadratic fit to the electrostatic energy;
2. add the EID additional analytical correction term, which removes the residual PBC-induced potential-offset contribution.

This repository therefore provides both the fitting workflow and a minimal script for the additional EID correction term.

Installation

git clone https://github.com/ZheLi-Lab/The_EID_charge_correction_method.git
cd The_EID_charge_correction_method
pip install numpy pandas scipy matplotlib pymbar alchemlyb pillow

Dependencies for the fitting workflow:

• Python 3.7+
• NumPy
• Pandas
• SciPy
• Matplotlib
• pymbar
• alchemlyb
• Pillow

Optional dependencies:

• OpenMM Used to generate simulation energy output files, but not required for the EID post-processing scripts in this repository.
• PyMOL Only needed for the optional visualization helper in AlchemConvTools/src/common_tools/drawgroup.py.

Usage

1. Prepare electrostatic energy data

Collect the potential-energy data from the electrostatic annihilation windows. The example files in example/ show the expected file layout.

2. Perform multi-window quadratic fitting and extract the linear term

python EID_charge_correction_analysis.py

This script fits the electrostatic energy to U = A * lambda^2 + B * lambda + C for each frame and writes the linearly separated PBC dataset derived from the fitted B term.

3. Compute the linearly separated PBC free energy

python AlchemConvTools/one_end_fe_aly.py -i example/input_aly.txt

This step applies BAR to the linearly separated dataset and yields the linearly separated PBC contribution.

4. Add the residual offset-removal analytical correction

python eid_additional_correction.py \
  --ligand-charge -1 \
  --environment-charge 0 \
  --box-length 60 \
  --system-name ligand_in_complex

The script writes a one-row CSV file named eid_additional_correction.csv by default and prints the correction in both kJ/mol and kcal/mol. In the standard EID setup, environment_charge should be zero. The final EID free energy is obtained by adding this term to the linearly separated PBC contribution from step 3.

If you want the binding-level additional correction directly, you can provide the complex-side and ligand-side box lengths together:

python eid_additional_correction.py \
  --ligand-charge -1 \
  --complex-box-length 60 \
  --ligand-box-length 40 \
  --system-name example_binding

In this mode, the output CSV contains the complex-side term, the ligand-side term, and the final binding-side difference for the additional correction.

Example input_aly.txt

[Basic_settings]
simulation_software = openmm
file_directory = "."
file_prefix = 'state_ΔUs'
file_suffix = '.csv'
subsample = False
fraction = 1.0
output_csv_filename = 'free_ene.csv'
energy_unit = 'kcal/mol'
calculation_windows = all
std_mode = bar_std
plot_du = False