BN-BacArena

Bayesian Network implementation in BacArena. This algorithm allows the dynamic modelling and simulation of bacterial communities considering the microbe-microbe and microbe-nutrient interactions. For more information, please refer to [paper]

The contents of BN-BacArena can be downloaded by cloning this repository.

git clone https://github.com/PlanesLab/BN-BacArena.git

Contents

• BN-Pipeline folder contains the example code required to generate the Bayesian network and extract the relevant microbe-microbe and nutrient-microbe interactions and their coefficients.
• R and MATLAB folders contain the source code required to launch BN-BacArena algorithm in both programming languages respectively.
• test folder contains the necessary data to launch BN-BacArena test scripts in both programming languages.
• The scripts testR.R and testMat.m allow to automatically run BN-BacArena algorithm in R and MATLAB respectively using data located at test folder.

Requirements

Specific libraries are required to generate the Bayesian network information.

• R: Please install openxlsx, dplyr, mclust and bnlearn libraries prior running BN_pipeline.R script. Remember that R >= 3.4.0 is required.

Prior installation of software/library is required to run BN-BacArena and/or test scripts.

• R: Please install BacArena and openxlsx libraries prior running BN-BacArena.
• MATLAB: The installation of The COBRA Toolbox is required to run BN-BacArena. More information available here.

Notes

• Note that random seeds are different between R and MATLAB languages. Therefore, BN-BacArena results could be different between them even if the same seed is applied.
• Using coefficient matrices is not a requirement for employing the BN-BacArena code (in both R and MATLAB). By doing so, users can still achieve the identical results provided by the original BacArena code. Accordingly, cell type and nutrient coefficient matrices can be applied independently.
• To use your own cell type and nutrient coefficient matrices, please remember that the dimensions of the matrices are cxc and cxn respectively, where c refers to the different cell types introduced in the arena and n to the nutrients of the culture media. The order of the rows and columns of the coefficient matrices should be equal to the order of the cell types and nutrients introduced in the arena.
• In order to use a nutrient coefficient matrix in R, it is mandatory to add the argument addAnyway = TRUE in the addSubs function of BacArena.

arenaWithMedia = addSubs(arena, ..., addAnyway = TRUE)