Eco-evolutionary dynamics of anthelmintic resistance in soil-transmitted helminths
Data files
Jan 02, 2024 version files 8.71 MB
-
50grid_host_v2_1_res.csv
30 KB
-
50grid_host_v2.csv
1.70 MB
-
50grid_hostsubset_res3.csv
30 KB
-
50grid_hostsubset.csv
1.70 MB
-
50grid_parasite_mul_res.csv
30 KB
-
50grid_parasite_mul.csv
1.61 MB
-
50grid_parasite_res.csv
30 KB
-
50grid_parasite.csv
1.68 MB
-
50grid_treatment_res3.csv
30 KB
-
50grid_treatment.csv
1.87 MB
-
README.md
3.72 KB
Abstract
Anthelmintic resistance (AR) of helminth parasites against the most widely available drugs is an ongoing concern for both human and livestock-infecting species. Indeed, there has been substantial evidence of AR in livestock but less in humans, which may be due to a variety of reasons. In this paper, we develop an eco-evolutionary model that couples the life cycle of these parasites with their underlying evolution in a single biallelic genetic locus that confers resistance to treatment drugs. We determine the critical treatment frequency needed to effectively eliminate the population, for a fixed drug efficacy (without evolution) and use this to classify three qualitative distinct behaviors of the eco-evolutionary model. Then, we describe how aspects of the life cycle influence which qualitative outcome is achieved and the spread of the resistance allele, comparing across human- and livestock- infecting species. For all but one species, we find that lower fecundity rates and lower contact rates speed the spread of resistance, while lower larval death slows it down. The life cycle parameters of Ancylostoma duodenale and Ostertagia circumcincta are associated with the fastest and slowest spread of resistance, respectively. We discuss the mechanistic reason for these results.
README: Eco-evolutionary dynamics of anthelmintic resistance in soil-transmitted helminths
https://doi.org/10.5061/dryad.5hqbzkhcn
These data are created as a output of an eco-evolutionary model detailed in the manuscript. We ran a number of different model parameterizations and report the ecological dynamics (growth, rescue, or decline coded as 1, 2, and 3). Both Mathematica and R code include the same eco-evolutionary model and includes parameter values for all six helminth species.
Description of the data and file structure
File list:
main_figures.R
SI_mathematica.nb
SI_mathematica.pdf
50grid_treatment.csv
50grid_treatment_res3.csv
50grid_parasite.csv
50grid_parasite_res.csv
50grid_parasite_mul.csv
50grid_parasite_mul_res.csv
50grid_host_v2.csv
50grid_host_v2_1_res.csv
50grid_host_subset.csv
50grid_host_subset_res3.csv
There are two sets of code. The Mathematica code (provided as SI_mathematica.nb and SI_mathematica.pdf) reproduces Figure 2 of the critical time between pulses as it depends on drug efficacy. The R code (main_figures.R) reproduces figures 3-8 and supplemental figures. Figures 3-5 require the datasets with "50grid" in the file name, which have been produced using the Mathematica code to categorize the ecological dynamics where the output is stored with the "res" suffix in a single column (growth, rescue, or decline coded as 1, 2, and 3 respectively).
The variables in the "50grid" datasets refer to the 10 parameters in the model. These variables are:
p = coverage of the population, i.e. the fraction of the population treated with drugs.
freq = the frequency of drug treatment in number of times per year.
mu_a = adult worm death rate in deaths per day.
lambda_eff = reproduction rate in eggs per day.
N = host population density in hosts per m^2.
beta_eff = contact rate per host in m^2 per day.
mu_l = larval death rate in deaths per day
mu_h = host death rate in deaths per day.
mu_BB = drug efficacy against worms of genotype BB, i.e. the fraction of worms that don't survive treatment, ranging from 0 to 1.
mu_AB = drug efficacy against worms of genotype AB, i.e. the fraction of worms that don't survive treatment, ranging from 0 to 1.
mu_AA = drug efficacy against worms of genotype AA, i.e. the fraction of worms that don't survive treatment, ranging from 0 to 1.
In "50grid_treatment.csv" file the values of frequency (freq) and coverage (p) are manipulated. The results are in "50grid_treatment_res3.csv".
In "50grid_parasite.csv" file the values of adult death rate (mu_a) and egg production rate (lambda_eff) are manipulated. The results are in "50grid_parasite_res.csv".
In "50grid_parasite_mul.csv" file the values of larval death rate (mu_l) and egg production rate (lambda_eff) are manipulated. The results are in "50grid_parasite_mul_res.csv".
In "50grid_host_v2.csv" file the values of contact rate (beta_eff) and host density (N) are manipulated. The results are in "50grid_host_v2_1_res.csv".
In "50grid_host_subset.csv" file the values of contact rate (beta_eff) and host density (N) are manipulated, but over a smaller range of contact rates. The results are in "50grid_host_subset_res3.csv".
The additional parameters host death rate (mu-h), drug susceptibility of genotype BB (mu_BB), death rate of genotype AB (mu_AB), and death rate of genotype AA (mu_AA) are held constant.
Sharing/Access information
All data was produced from our model.
Code/Software
We suggest running the Mathematica code first and then the R code. The R code will call the "grid" datasets in the "Grids" folder to create figures 3-5. R version 4.1.1 was used and the following packages were loaded: cowplot, deSolve, ggplot2, ggpubr, gridExtra, viridis.