Installation

Installation#

Dependencies#

  • Linux operating system

  • g++ >= 4.8

  • CUDA toolkit >= 6

  • ARBD simulation engine

  • Python >= 3.5 (some users report problems installing mdanalysis; miniconda or conda are recommended) * numpy >= 1.14 * scipy >= 1.1 * mdanalysis >= 0.18 * cadnano >= 2.5.2.1 * appdirs >= 1.4 * pandas >= 1.0 * scadnano >= 0.1 (optional) * oxDNA (optional)

Installation Steps#

First make sure you have the cuda-toolkit installed. First download ARBD through a web browser. The following script can be used as a guide to install mrdna.

# Customize the following
BUILD_DIR=/path/to/build-dir
export CUDA_PATH=/path/to/cuda-toolkit

# Unpack and build arbd
cd $BUILD_DIR
mv ~/Downloads/arbd*.tar.gz .
tar -xzf arbd*.tar.gz
cd arbd
cmake -S . -B build && (
  cd build
  make -j
)
cd ../../

## Possibly setup python dependencies (you may want to set up a virtual environement)
# conda install "numpy>=1.14" "scipy>=1.1" "appdirs>=1.4"
# conda install -c conda-forge "mdanalysis>=0.18" "pandas>=1.0"
# pip3 install "cadnano>=2.5.2.1" "scadnano>=0.1"

git clone https://gitlab.engr.illinois.edu/tbgl/tools/mrdna
cd mrdna
pip3 install .
cd ..

# Finally update your bashrc so that your shell can find the ARBD binary
echo "PATH=$BUILD_DIR/arbd/src:\$PATH" >> ~/.bashrc
source ~/.bashrc

Please report issues to Chris Maffeo.