Transmission: A tool for inferring symbiont transmission mode from metagenomic data

Installation

Install r package abc in the usual way, then install Transmission with

pip install transmission-popgen

Generating simulated summary statistics

This is a big job and I find it helpful to farm this job out to a computing cluster. The main program is written for interactive use in a notebook or shell, so to that end, the cli tool transmission-priorgen is included with transmission. Use transmission-priorgen --help for details on available options.

Natively

For example:

transmission-priorgen -n 10 -d 5 -M 2 -s 100 \
    -p '{"eta": (0, 0.1), "tau":(1, 1), "rho"(10, 10)}' \
    --h_opts '{"bias": False}' outfile.pickle

Using the Docker image

There is a docker image available to use the command line tools as well as to launch a Jupyter notebook server. It can be found here. The most recent push to master is tagged latest, however it is probably desirable to use explicit version numbers for reproducibility.

To generate priors

docker run --rm -v </path/to/host/directory>:/home/jovyan/work \
    mpjuers/transmission:<version> \
    transmission-priorgen [options] /home/jovyan/work/<outfile.pickle>

--rm removes containers you are finished with while -v ("volume") binds a directory on the host machine to one on the container. should be prefixed by 'v', e.g. v0.0.3

If you are working on a compute cluster, you might have access to Singularity rather than Docker. The transition is straightforward:

singularity exec --contain -B </path/to/host/directory>:/home/jovyan/work \
    docker://mpjuers/transmission:<version> \
    transmission-priorgen [options] /home/jovyan/work/<outfile.pickle>

You may or may not need the --contain flag; I needed because I had some locally installed python modules that conflicted with those in the container.

Using the Docker image to run a Jupyter notebook server.

I have had some difficulty compiling rpy2 on my platform (MacOS Mojave). This may or may not still be the case, but if you wish to avoid any problems associated with setting up transmission on your machine, it is possible to run a Jupyter server from the Docker image. First, launch Docker on your machine and then run, e.g.,

docker run --rm -p 8888:8888 -e JUPYTER_ENABLE_LAB=yes \
    -v </path/to/local/dir>a:/home/jovyan/work mpjuers/transmission:<version>

Omit -e JUPYTER_ENEABLE_LAB=yes if you are going to use a vanilla notebook rather than Jupyter Lab. Then you can just open <hostname>:8888/ in a browser and enter the token that appears in the terminal and anything saved to ~/work on the container will appear in </path/to/local/dir> (and vice-versa). For more information on running Jupyter Docker containers, go to [https://github.com/jupyter/docker-stacks].