Repository containing dynamical models for the Microbiome Invasion and Transmission of plasmid-mediated Antimicrobial Resistance (MITAR) project.
To use this repository, download it as a ZIP file (use the green 'Code' button
and choose 'Download ZIP') and unzip the folder. You can move the folder to your
preferred location, but the R-scripts should be in the same folder as the
R-project file MITAR.Rproj.
If you have not yet installed R, download it from CRAN and install it. You might want to install an integrated development environment for R, such as RStudio. Brief instructions on installing R, RStudio, and R-packages are given below. More extensive information is available from the InstallPkgs GitHub repository.
After R has been installed, install the required R-packages: run R as
administrator and run the script installpkgs_mitar.R to install the required
R-packages. During the installation, you might be asked to close R and restart
R. The required packages are:
deSolve and
rootSolve to integrate ODEs
and assess equilibria,
TruncatedNormal for random
number generation from truncated distributions,
dplyr and
tidyr for data handling, and
ggplot2,
ggrepel,
scales and
cowplot for plotting.
To run the analyses, open the R-project file MITAR.Rproj and run one of the
R-scripts following the instructions in the next section 'File overview'.
The R-script to install the R-packages required by the other scripts (see the section Workflow above):
- installpkgs_mitar.R
R-scripts to compare pair-formation models of conjugation with bulk-conjugation models, used in Alderliesten JB, Zwart MP, de Visser JAGM, Stegeman A, Fischer EAJ. 2022. Second compartment widens plasmid invasion conditions: two-compartment pair-formation model of conjugation in the gut. Journal of Theoretical Biology 533:110937.
- pairformation.R
- pairformation2comp.R
The first script is used to model conjugation in a single compartment, the second script to model conjugation in two compartments with migration between them. Select one parameter set from the section 'Parameter values' at a time, run the section 'Run simulations', and run the appropriate output section. Results of the simulations are saved as Comma Separated Values (.csv) files. Graphs are created and, if the variabe 'saveplots' in the section 'Plotting and simulation options' is set to TRUE, saved as .png-files. If the variable 'plotdataapproxbulk' is set to TRUE, the data used to approximate the bulk-conjugation rates is plotted.
R-scripts and MATLAB-script for multispecies models of conjugation, used in Alderliesten JB, Zwart MP, de Visser JAGM, Stegeman A, Fischer EAJ. In prep. The effects of ecological interactions and distinct conjugation rates on the invasion of a conjugative plasmid in bacterial communities.
These scripts use a generalised Lotka-Volterra model elaborated with plasmid-bearing populations and conjugation to investigate how the 7607 richness, evenness, and diversity of a plasmid-free microbiome influence the ability of plasmid-bearing bacteria to invade it (results for plasmid-free bacteria as invaders are also included in the results of the simulations). Throughout, the interspecies conjugation rates involving the invading species (i.e., the initially plasmid-bearing species) are either equal to the other interspecies conjugation rates or are thousandfold lower.
- multispecies.R: simulate invasion through a species that is already present at the plasmid-free equilibrium. Interaction coefficients are drawn from distributions, species abundandances are specified by species abundance models, and growth rates are calculated from the interactions and abundances to obtain an equilibrium.
- multispecies_altgrow.R: similar to multispecies.R, but here interaction coefficients and growth rates are drawn from distributions and then species abundances are calculated from them to obtain an equilibrium.
- multispecies_analytic.m: MATLAB-script to obtain an analytical expressions for the eigenvalues and derive (in)stability criteria in the two-species case.
- multispecies_bifur.R: R-script to create bifurcation-like plots showing how the intraspecies conjugation rate of the initially plasmid-bearing species and fitness costs of bearing a plasmid influence the epidemiological stability of the plasmid-free equilibrium (Figures 2 and S6).
- multispecies_distr.R: R-script to create examples of probability densities for the distributions of ecological interaction coefficients (Figure S1).
- multispecies_funcs.R: R-script containing the functions used by the other R-scripts.
- multispecies_pinnew.R: similar to multispecies.R, but here invasion through a species that was not yet present at the plasmid-free equilibrium is simulated.
- multispecies_pinnew_bifur.R: similar to multispecies_bifur.R, but here invasion through a species that was not yet present at the plasmid-free equilibrium is simulated.
To run the simulations, choose one of the R-scripts and run it from the top to load the required packages, select the basis parameter set from the section 'Settings and defining parameter space', run the simulations and plot the results. Results of the simulations are saved as R-data (.RDS) files. Graphs are created and, if the variabe 'saveplots' in the section 'Settings and defining parameter space' is set to TRUE, saved as .png-files. If the variable 'saveplotconjovertime' is set to TRUE, plots showing the dynamics over time are created and saved.
Changes to the settings allow to model further scenarios. For example, the initial total abundance, fitness costs of bearing a plasmid, and interspecies conjugation rate can be changed. In addition, the default scenario where plasmid-bearing bacteria replace bacteria of the most-abundant species can be changed to let them replace bacteria of the least-abundant species by changing variable 'PReplMostAbun' from TRUE (the default) to FALSE. For scripts that simulate invasion through a new species, the new species that introduces the plasmid grows slower than, equally fast as (default), or faster than the already-present species, depending on the value of variable 'newgrowthratecode' (1, 2, or 3, respectively).