Code to replicate experiments from Compositional Lotka-Volterra describes microbial dynamics in the simplex.

Environment

All experiments were performed using Python 3.5.2. Packages, along with version numbers, can be found in requirements.txt. Most (if not all) of the code should run on any version of Python 3, and most versions of the packages in requirements.txt.

Generating Figures

The following scripts generate the panels for the main results of the paper.

Figure 3

plot_concentrations.py
bucci_cdiff_lokta_volterra_comparison.py
bucci_diet_lokta_volterra_comparison.py
stein_lokta_volterra_comparison.py

Figure 4

plot_prediction_results.py

Figure 5 and S9

stein_plot_parameters.py

Figure S5, S6 and S7

plot_prediction_trajectories.py

Rerunning Experiments

Each experiment is organized by the dataset on which it was performed. Datasets are labeled by FIRST_AUTHOR (e.g. Bucci, Stein).

Model Comparison

The following scripts run the model comparison experiments:

bucci_cdiff_prediction_experiments.py
bucci_diet_prediction_experiments.py
stein_prediction_experiments.py

Simulated Data

The script simulation_bucci_clv.py is responsible for generating simulated data and performing the comparison between elastic net and ridge regression. It uses a command line interface to set simulation parameters, an example of which can be found in simulation_bucci_clv.sh. The script plot_simulation_bucci_clv.py plots the simulation results saved in tmp_sim.

Data and Preprocessing

Data and preprocessing steps can be found under data/FIRST_AUTHOR.