Hess

Hess is a toolkit for computational chemistry. This public version of the toolkit provides functionality for empirical and semi-empirical molecular docking.

Prerequisities

• GNU/Linux operating system.
• Development tools:

  sudo apt-get install g++ make libz-dev libeigen3-dev makedepf90 libstdc++-12-dev
  

• Intel OneAPI toolkit with icx and ifort compilers (if you need Hess SE):

  sudo apt-get install intel-oneapi-mkl-devel-2022.2.1 intel-oneapi-compiler-fortran-2022.2.1 intel-oneapi-compiler-dpcpp-cpp-2022.2.1 
  

Build

With Hess SE

source /opt/intel/oneapi/setvars.sh

./setup.sh

Indigo library will be downloaded in the process.

Without Hess SE

./setup-no-se.sh

Indigo library will be downloaded in the process.

Empirical docking (includes preparation)

Usage

hess-empirical-docking [options]

Input options

PDB and Molfile formats are accepted.

• -r, --receptor arg: path to the receptor molecule
• -l, --ligand arg: path to the ligand molecule
• --autobox_ligand arg: ligand to use for autobox calculation

Box options

Units are Angstroms.

• --center_x arg: X coordinate of the center
• --center_x arg: Y coordinate of the center
• --center_x arg: Z coordinate of the center
• --size_x arg: size in the X dimension
• --size_y arg: size in the Y dimension
• --size_z arg: size in the Z dimension

Minimization algorithm options

• --optimize arg: global search algorithm variant (mc_metropolis for Monte Carlo with Metropolis criterion or mc for Monte Carlo optimisation)
• --top arg: number of top scored poses stred in the output
• --depth arg: search depth (number of LBFGS local search runs)
• --granularity arg: grid granularity (the smaller, the better the approximation)
• --score_only: score provided ligand pose
• --grid_deriv: use a grid with precalculated gradients

Output

• -o arg: output file name (sdf)

Other

• --seed arg: explicit random seed

Example run

Go to the hess-empirical-docking folder.

./dist/Release/GNU-Linux/hess-empirical-docking -r ../data/protein_example.pdb -l ../data/ligand_example.pdb --autobox_ligand ../data/crystal_example.pdb --number_of_iterations 16 -o results.sdf

Preparation

Usage

hess-preparation <path to input pdb file> <path to output molfile>

Example

Go to the hess-preparation folder.

./dist/Release/GNU-Linux/hess-preparation ../data/molecule_example.pdb prepared_molecule.mol

SE program

SE program calculates the energy of a molecule. The program does not include the preparation of the molecule.

Usage

For non-optimization run:

./hess-se-docking -p <path to input pdb file> -popt n -pout <path to output molfile>

For optimization run:

./hess-se-docking -p <path to input pdb file> -popt y -pout <path to output molfile>

Example

Without optimization

Go to the hess-se-docking folder and run the energy calculation for 6MDC complex:

./dist/Release/GNU-Linux/hess-se-docking -p ../data/test-se/6mdc/6mdc_c_h.pdb -popt n -pout ../data/test-se/6mdc/6mdc_c_h_out.mol

And then for the ligand and the protein separately:

./dist/Release/GNU-Linux/hess-se-docking -p ../data/test-se/6mdc/6mdc_l_h.pdb -popt n -pout ../data/test-se/6mdc/6mdc_l_h_out.mol

./dist/Release/GNU-Linux/hess-se-docking -p ../data/test-se/6mdc/6mdc_p_h.pdb -popt n -pout ../data/test-se/6mdc/6mdc_p_h_out.mol

After each run, the energy will be printed out; to obtain the binding energy, the energy of the protein and ligand should be subtracted from the energy of the complex.

With optimization

Here is a similar example for the 6S9B complex, but now with optimization enabled. Optimizing the protein and complex will take considerable time.

./dist/Release/GNU-Linux/hess-se-docking -p ../data/test-se/6s9b/6s9b_c_h.pdb -popt y -pout ../data/test-se/6s9b/6s9b_c_h_out.mol

./dist/Release/GNU-Linux/hess-se-docking -p ../data/test-se/6s9b/6s9b_l_h.pdb -popt y -pout ../data/test-se/6s9b/6s9b_l_h_out.mol

./dist/Release/GNU-Linux/hess-se-docking -p ../data/test-se/6s9b/6s9b_p_h.pdb -popt y -pout ../data/test-se/6s9b/6s9b_p_h_out.mol