MDPbiomeGEM is a combination of the published system MDPbiome and Genome-scale Metabolic models (GEMs). Multiple microbial community dynamics under perturbations timeseries are simulated by MMODES, followed by prediction of the interventions required to direct the engineering of such a microbial community into a desirable state with MDPbiome.

MDPbiomeGEM identifies different microbial community states arising in the simulated data, and the perturbation-caused transitions between states. It then computes a recommendation consisting of directed interventions that will move microbial communities from a disease or low-performance state to a more desirable state.

MDPbiome is a combination of different technologies that can be run in three docker containers (recommended). Alternatively, you could download and install the pre-requisites individual software by yourself.

Once you have the docker daemon running, download the different docker images from the cloud:

docker pull carrascomj/mmodes
docker pull beatrizgj/mmodes-to-mdpbiome
docker pull beatrizgj/mdpbiome

Before run the docker images, you need to create a folder in the host (for example, $HOME/DockerOutputMDPbiomeGEM), to exchange the data between different docker images, where the output/input data will be copied. If you do not create the folder before run the docker, it would be created automatically with NO permissions to write them.

To design and code your simulations in a run_MyCommunity_simulations.py, following the MMODES documentation (https://mmodes.readthedocs.io/) and/or the examples of run_BifFae_simulations.py and run_Atrazine_simulations.py, saving in the host folder to share (for example: $HOME/DockerOutputMDPbiomeGEM)

Run simulations:

You can install MMODES locally in your host, following the instructions at https://github.com/carrascomj/mmodes or you can run the MMODES docker image:

sudo docker run -it --volume "$HOME/DockerOutputMDPbiomeGEM":/home/cleanapp carrascomj/mmodes

Copy your designed simulation file (run_MyCommunity_simulations.py) in "$HOME/DockerOutputMDPbiomeGEM" and your GEMs in the right pathway (ex. "$HOME/DockerOutputMDPbiomeGEM/draft/ModelsInput/").

As an example, you could copy the files from https://github.com/beatrizgj/MDPbiomeGEM/tree/master/Src/DockerMMODES, or inside the docker container from /app directory.

Inside the docker (or in the bash terminal), run the simulations with:

python3 run_Atrazine_simulations.py <num_processors> <num_simulations>

For example, running in background, for example:

nohup python3 run_Atrazine_simulations.py 6 10 >& output_runMMODES.txt &

To run the complementary scripts to collect all the simulations:

python3 to_results.py <directory_with_outputfiles>

To move the biomass*.tsv files to a folder that should be accesible for the next steps. Following with the example '$HOME/DockerOutputMDPbiomeGEM/CommunityExample/'

sudo docker run -it --rm -p 8888:8888 -v "$HOME/DockerOutputMDPbiomeGEM":/home/jovyan/work beatrizgj/mmodes-to-mdpbiome

Click or copy the last URL to open jupyter in the web browser

Inside jupyter:

a.- Go to MMODEStoMDPbiome folder, open and run step by step runMMODEStoMDPbiome.ipynb, personalized the steps requiring USER-INTERACTION

b.- Then, you need to move the folder with the results (for example 'CommunityExample') to /work folder, that is the folder shared with your host

Close the mmodes-to-mdpbiome docker container (because the next one runs in the same port)

sudo docker run -it --rm -p 8888:8888 -v "$HOME/DockerOutputMDPbiomeGEM":/home/jovyan/work beatrizgj/mdpbiome

Click or copy the last URL to open jupyter in the web browser

Inside jupyter:

a.- Go to work/ (that is the folder shared with your host) and to move the folder with the results from the previous step (for example 'CommunityExample') to /MDPbiome folder. If there are a folder with the same name in MDPbiome/, you should remove or rename it before

b.- Go to MDPbiome folder, open and run step by step config_MDPbiome_template.ipynb, personalized the steps requiring USER-INTERACTION

c.- Then, you need to move the folder with the results (for example 'CommunityExample') to /work folder, to preserve the result files when you close the docker

The input and output data files used in the current MDPbiomeGEM case studies are available at Zenodo: http://doi.org/10.5281/zenodo.3667601