TruSD 0.0.1

TruSD co-infers selection coefficients and genetic drift from allele trajectories using a maximum-likelihood framework.

TruSD

Trajectories under Selection and Drift is an implementation of a method that co-infers selection coefficients and genetic drift from allele trajectories using a maximum-likelihood framework.

If you find the software useful in your research please cite this package.

Installation

TruSD needs Python 3.6 or newer. Python 2 is not supported!

Currently, the best way to install TruSD is to use a virtual environment and install TruSD from source.

To setup a virtual environment, go to some folder where you want to have your virtual environment. This could be your home folder. The following command will create the virtual environment in the folder trusd_env. You can call the environment something else, but then you have to adapt the name also in the examples further below.

python3 -m venv trusd_env

This will take some seconds. If you do not have venv installed, you can install it running apt-get install python3-venv. After creating the virtual environment, you can activate it:

source trusd_env/bin/activate

Your command line prompt should change to reflect the activated environment. Now, you can install TruSD into that environment. You can also install other packages that you might think are useful for your work. Make sure that pip is uprade to the latest version (first line in the following).

pip3 install --upgrade pip
git clone https://github.com/mathiasbockwoldt/TruSD.git
cd TruSD
pip3 install -e .

If you want to leave the environment again, run deactivate.

You should now have a folder structure like this:

├── trusd_env
│   ├── bin
│   │   └── activate
│   └── [other files of your virtual environment]
└── TruSD
    ├── min_working_example.py
    └── [other files of TruSD]

Update

If you later want to update TruSD, go to the TruSD folder (not the trusd_env folder) and run:

git pull

Running TruSD

If you installed TruSD in a virtual environment, you have to activate it, before you can run TruSD. The command line prompt should show you, whether it is active.

source trusd_env/bin/activate

Please note, that you are not in the right path, you will have to give a full absolute or relative path:

source /path/to/trusd_env/bin/activate
# or, e.g., if you are in the TruSD folder:
# source ../trusd_env/bin/activate

Now you should be able to run TruSD and, for example, get some help.

trusd --help

You can also import the trusd module from your Python 3 script:

# inside python3 script or interactive prompt
import trusd
help(trusd.trusd) # gives package content

You can start an example run with

python3 min_working_example.py

The min_working_example.py is documented to explain the basic steps. It uses traj_example.txt as input and produces outfile.txt (the results), outfile.json (the metadata) and outfile.pdf (the plot).

To get out of the virtual environment, run deactivate in your shell.

If you want to run TruSD from a virtual environment on a cluster, please make the step source /path/to/trusd_env/bin/activate part of your job script.