QMFlows

See http://scm-nv.github.io/qmflows/ for tutorials and documentation.

Motivation

Research on modern computational quantum chemistry relies on a set of computational tools to carry out calculations. The complexity of the calculations usually requires intercommunication between the aforementioned tools, such communication is usually done through shell scripts that try to automate input/output actions like: launching the computations in a cluster, reading the resulting output and feeding the relevant numerical result to another program. Such scripts are difficult to maintain and extend, requiring a significant programming expertise to work with them. Being then desirable a set of automatic and extensible tools that allows to perform complex simulations in heterogeneous hardware platforms.

This library tackles the construction and efficient execution of computational chemistry workflows. This allows computational chemists to use the emerging massively parallel compute environments in an easy manner and focus on interpretation of scientific data rather than on tedious job submission procedures and manual data processing.

Description

This library consists of a set of modules written in Python3 to automate the following tasks:

1. Input generation.
2. Handle tasks dependencies (Noodles).
3. Advanced molecular manipulation capabilities with (rdkit).
4. Numerical data storage and manipulation (HDF5).
5. Jobs failure detection and recovery.
6. Distribution in heterogeneous hardware platforms.

Tutorial and Examples

A tutorial written as a jupyter-notebook is available from: tutorial-qmflows. You can also access direclty more advanced examples.

Installation

• Download miniconda for python3: miniconda (also you can install the complete anaconda version).
• Install according to: installConda.
• Reopen terminal (or type source ~/.bashrc).
• Create a new virtual environment using the following commands:
  • conda create -n qmflows
• Activate the new virtual environment
  • source activate qmflows

To exit the virtual environment type source deactivate.

Dependencies installation

Using the conda environment the following packages should be installed:

• install rdkit using the following command:
  • conda install -y -q --name qmflows -c https://conda.anaconda.org/rdkit rdkit
• install HDF5 using conda:
  • conda install -y -q --name qmflows -c anaconda h5py

Package installation

• Type in your terminal,

  source activate qmflows

• Then

  pip install https://github.com/SCM-NV/qmflows/tarball/master#egg=qmflows

Now you are ready to use qmflows.

Notes:

• Once the libraries and the virtual environment are installed, you only need to type source activate qmflows each time that you want to use the software.