1. Installation

At this step, you may have to chose between either a blue or a red pill 💊.

With the blue pill, simply install the bayescode conda package from the channel bioconda. Or chose the red pill and install the requirements and compile the C++ code for BayesCode in your Debian based OS.

Blue pill is preferred is you want minimal conflict with your local system. Red pill method is preferred if you plan to tinker with the code. The two pills are mutually not exclusive, no overdose had ever been observed (though no statistical study had been performed).

1. Blue pill with conda

The precompiled binaries for nodemutsel, readnodemutsel, mutselomega, readmutselomega and python scripts are available as the bayescode conda package from the channel bioconda (anaconda.org/bioconda/bayescode).

conda install -c conda-forge -c bioconda bayescode

You can then go to the section Format your data.

2. Red pill and compile code

Requirements to compile BayesCode are a C++ compiler (clang or g++), make and cmake:

sudo apt install make cmake clang

To get and build BayesCode:

git clone https://github.com/ThibaultLatrille/bayescode
cd bayescode
make tiny

To check that everything ran well, look into the bin folder to see if executables are present:

• nodemutsel
• readnodemutsel
• mutselomega
• readmutselomega

To update to the latest version of BayesCode simply run:

git pull
make clean
make tiny

Optional: to compile the whole BayesCode suite, a compatible MPI compiler is required

sudo apt install openmpi-bin openmpi-common libopenmpi-dev
make release