quanformer: quantum mechanical analysis of molecular conformers

quanformer is a Python-based pipeline for generating conformers, preparing quantum mechanical (QM) calculations, and processing QM results for a set of input molecules.

For each molecule, conformers are generated and optimized using a molecular mechanics force field. Input files for QM calculations are then prepared for geometry optimizations, single point energy (SPE) calculations, or Hessian calculations. The user specifies any QM method and basis set that is supported in the QM software package (Psi4 or Turbomole). With completed calculations, the pipeline extracts final energies and geometries and collects job-related details such as calculation time and number of optimization steps. Analysis scripts are provided for comparing different methods in the lens of conformer energies and calculation times via easily-interpreted plots. This pipeline was tested to robustly process hundreds of conformers per molecule and hundreds of molecules, though it can likely handle more.

This code is open source on Github.

Installing quanformer

Anaconda

1. Install Miniconda (or Anaconda) with Python 3.6 (version 3.7 is not yet tested).

2. Create a conda environment for quanformer:

   conda create --name quanformer python=3.6 matplotlib scipy
   conda activate quanformer
   

3. Obtain OpenEye, Psi4, and Psi4 dependencies:

   conda install -c openeye openeye-toolkits
   conda install -c psi4 psi4 dftd3 gcp [gpu_dfcc]
   

4. Check installations.

   1. Psi4:

      psi4 --test
      

   2. OpenEye (in interactive Python session):

      import openeye.oechem as oechem
      mol = oechem.OEMol()
      

Quanformer

Install quanformer from source by the following:

1. Clone the repository from Github:

   git clone https://github.com/MobleyLab/quanformer
   cd quanformer
   

2. Install:

   python setup.py install
   

   or use pip for a local install (editable installation):

   pip install -e .
   

quanformer API

coming soon

Utility functions

coming soon

Examples

Example files are provided on Github.

Remember to activate your conda environment containing quanformer:

conda activate quanformer

Basic usage

Set up structures and prepare calculations:

import quanformer.pipeline as qp

qp.setup_conformers('two_alkanes.smi')
qp.setup_calculations('two_alkanes-200.sdf','mp2','def2-sv(p)')

Your directory structure should look something like this:

.
├── AlkEthOH_c1008
│   ├── 1
│   │   └── input.dat
│   ├── 2
│   │   └── input.dat
│   └── 3
│       └── input.dat
├── AlkEthOH_c312
│   ├── 1
│   │   └── input.dat
│   ├── 2
│   │   └── input.dat
│   └── 3
│       └── input.dat
├── numConfs.txt
├── two_alkanes-200.sdf
└── two_alkanes.sdf

After running calculations from provided input files, collect results:

qp.process_results('two_alkanes-200.sdf')

Utilities

quanformer comes with a number of utility functions. These can be used as shown here:

import quanformer.utils as qu
qu.convert_extension('two_alkanes-200.sdf','two_alkanes-200.mol2')