This repository contains a working set of examples, current documentation and details of compiling and running pyCHARMM. The materials here are part of a growing repository for examples and tutorials on using pyCHARMM. As we evolve the program, expand its example base and run new workshops based on pyCHARMM, we anticiapte this repository will grow.
We note that pyCHARMM is now part of CHARMM distribution - free of charge to academic and government/non-profit labs. The CHARMM/pyCHARMM source is available through academic licensing at academiccharmm.org.
Date: June 20-24, 2022
Time: 4:00 - 5:30 PM EDT
Location: Chem 1706
For those who do not have access to gollum or would like to set up their own conda environment for pycharmm, clone this repo and create the conda environment by
cd pyCHARMM-Workshop/
conda create -n pyCHARMM -c conda-forge -c schrodinger \
pymol rdkit jupyterlab openbabel openmm
conda activate pyCHARMMWe will utilize pyCHARMM from the current developers stream. If you are working on gollum you can link to the pyCHARMM version I compiled using (tsch):
export CHARMM_HOME=/home/brookscl/charmm/c47-dev-release/
export CHARMM_LIB_DIR=$CHARMM_HOME/install-pycharmm-nompi/libInstall pyCHARMM in the conda environment using pip
pip install $CHARMM_HOME/tool/pycharmmIf you are building pyCHARMM in your local environment (assuming Linux, gcc (I use v7.3), NVIDIA drivers (I use cuda v10.0)/gpu access with OpenMM already installed), in the CHARMM directory:
mkdir -p build_charmm; cd build_charmm
../configure --with-blade --with-fftdock --without-mpi \
--as-library -p ../install-pycharmm-nompi
make -j 2 -C installFollowing this you will need to set the environment variable CHARMM_LIB_DIR as noted above to point to your install library directory, and install the pyCHARMM modules using pip.
pip install ../tool/pycharmm
export CHARMM_LIB_DIR=/path/to/charmm_dir/install-pycharmm-nompi/lib- MMTSB Toolset is utilized and should be installed .
- propKa will be utilized.
- pymol will be utilzied locally to view structures.
If you have pycharmm installed locally in the pyCHARMM virtual environment, you can start a jupyter-lab by typing jupyter lab in the same conda environment.
- FirstExample - build an alanine dipeptide and solvate it
- ProteinDynamics - set-up solvated protein system and run dynamics on it
- SimpleMPIExample - use mpi to compute alanine dipeptide energy surface and plot it
- AbsoluteSolvation - Distributed (mpi) calculation of the solvent half of the absolute solvation free energy using MBAR and BAR
-
Aladipeptide_HFBString_MPI - Construct minimum energy path for
$\phi,\psi$ conformational transition in alanine dipeptide using HFB-String approach and mpi task distribution -
ReplicaExchangeExample - Compute the free energy change for charging a residue in vacuum with replica exchange
$\lambda$ -windowing distributed simulations using mpi - CDOCKER_Tutorial - Tutorial on using CDOCKER pyCHARMM implementation for flexible ligand - rigid receptor and flexible ligand - flexible receptor docking
- DrudeExample - Simple example of setting-up and using Drude oscillator polarizable model
- NeuralNetExample - Use of TorchANI machine learned quantum chemistry potentials in CHARMM calculations through pyCHARMM
-
msld_py_prep_Tutorial - Tutorial on setting-up and using MS
$\lambda$ D free energy calculations with py_prep
Introduction to CHARMM, how to build pyCHARMM and first steps in setting up molecular simulations with pyCHARMM, selecting atoms and basic manipulations of the PSF using pyCHARMM.
Setting up solvated molecular dynamics simulations, using OpenMM, BLaDE GPU accelerated simulation engines for molecular dynamics.
Using pyCHARMM CDOCKER for rigid-receptor, flexible ligand and flexible-receptor, flexible ligand docking.
Free energy simulations i) absolute solvation free energy calculations; ii) setting-up MSλD with msld-py-prep and pyCHARMM.
Advanced topics including integration of python tools with pyCHARMM, utilizing machine-learned potentials in pyCHARMM, parallelism in python with MPI.
