Early epidemiological characteristics explain the chance of population-level virus persistence following spillover events.

This repository contains the data and code used in the PLOS Biology submission "Early epidemiological characteristics explain the chance of population-level virus persistence following spillover events." Below we describe 1) the contents of each uploaded datasheet 2) the contents of each uploaded R file 3) how figures in the publication were created. Data files File Description FCXexperiment.population.csv Contains RT-qPCR Ct values for each passage plate in the first experimental block (FCX). FCYexperiment.population.csv Contains RT-qPCR Ct values for each passage plate in the second experimental block (FCY). FCZexperiment.population.csv Contains RT-qPCR Ct values for each passage plate in the third experimental block (FCZ). FDIexperiment.population.csv Contains RT-qPCR Ct values for each passage plate in the fourth experimental block (FDI). FCXstrips.csv Contains the RT-qPCR Ct values for individual and groups of worms used to calculate infection prevalence and intensity for the first experimental block (FCX) FCYstrips.csv Contains the RT-qPCR Ct values for individual and groups of worms used to calculate infection prevalence and intensity for the seoond experimental block (FCY) FCZstrips.csv Contains the RT-qPCR Ct values for individual and groups of worms used to calculate infection prevalence and intensity for the third experimental block (FCZ) FDIstrips.csv Contains the RT-qPCR Ct values for individual and groups of worms used to calculate infection prevalence and intensity for the fourth experimental block (FDI) FCXshedding.csv Contains counts of glowing and non-glowing plates in shedding assay first block (FCX) FCYshedding.csv Contains counts of glowing and non-glowing plates in shedding assay second block (FCY) FCZshedding.csv Contains counts of glowing and non-glowing plates in shedding assay third block (FCZ) FDIshedding.csv Contains counts of glowing and non-glowing plates in shedding assay fourth block (FDI) Strainsandblocks.csv Contains information about which strains/thaw lines were used in each block. TCID50bythaw.csv Median tissue culture infectious dose (TCID50) values of our stock virus in each strain and thaw line calculated through a maximum likelihood analysis in "TCID50Analysis.R" TCID50combined.csv Median tissue culture infectious dose (TCID50) values of our stock virus in each strain (thaw lines combined when not significantly different) calculated through a maximum likelihood analysis in "TCID50Analysis.R" TCID50Assays.csv Raw data (Ct values) from TCID50 Assays. These data are used in the R code "TCID50Analysis.R" to produce the files "TCID50bythaw.csv" and "TCID50combined.csv". passage.ctvalues.csv Combined RT-qPCR values for virus RNA1 detection in passage plates across the experiment. This datasheet is wrangled in R code "SpilloverCharacteristicsDataPrep.R". ModelAnalysis.Data.csv Contains mechanistic model predictions alongside actual passage success in the experiment. This dataframe produces figure 4 and is derived from "spill.char.dataset2.csv" through the R code "PredictingPersistenceMechModel.R". spill.char.dataset1.csv Combined data on the duration of virus detection through passages with spillover characteristics data for each passage line. This datasheet is wrangled through R code "SpilloverCharacteristicsDataPrep.R." spill.char.dataset2.csv Combined data on the success of each passage with spillover characteristics data for each passage line. This datasheet is wrangled through R code "SpilloverCharacteristicsDataPrep.R." 2. Code Files File Description SpilloverCharacteristicsDataPrep.R This code takes the raw datafiles and produces the combined datasets "spill.char.dataset1.csv" and "spill.char.dataset2.csv", which are used in the correlative and mechanistic models (see R code below). Code to produce Figure 2 as well as many supplemental figures is contained in this file as well. Of particular interest, the code to calculate the maximum likelihood shedding and prevalence estimates is L170-233. Correcting the infection intensity to account for high infection prevalence is L258-297. Calculation of probability of passage is L488-499. TCID50Analysis.R This code takes the raw data from the median tissue culture infectious dose experiments and calculates a maximum likelihood TCID50 for each strain. It produces the datafiles "TCID50bythaw.csv" and "TCID50.combined.csv". PredictingPersistenceCorrelativeModels.R This code runs the correlative models described in the paper. It also produces figure 3. PredictingPersistenceMechModel.R This code runs the mechanistic model described in the paper as well as the model comparison analysis. It also produces figure 4. 3. Figures Figure 2 The dataset "passage.ctvalues.csv" contains the ct values across passages shown in Figure 2A. The dataset "spill.char.dataset1.csv" contains the data depicted in Figure 2B-E. Specifically, 2B is plotting column "best.p", 2C is plotting "best.y", 2C is plotting "median.ct", 2D is plotting "rel.susc" (and 95% confidence intervals "rel.susc.low" and "rel.susce.high", 2E is plotting "p.paper". The R code "SpilloverCharacteristics.DataPrep" contains code for datawrangling the raw data into "spill.char.dataset1.csv" and graphing. Figure 3 The dataset "spill.char.dataset1.csv" contains the data depicted in Figure 3A-D. This figure plots duration of virus persistence through passages ("trans.ability") against infection prevalence ("best.p"; 3A), shedding ability ("best.y"; B), infection intensity ("median.ct"; 3C), and relative susceptibility ("rel.susc"; 3D). The R code "SpilloverCharacteristics.DataPrep" contains code for datawrangling that raw data into "spill.char.dataset1.csv". The R code "PredictingPersistenceCorrelativeModels.R" contains the code for making this figure. Figure 4 The dataset "ModelAnalysis.Data.csv" contains the data depicted in Figure 4, which plots the predictions of the mechanistic model against the actual passage success in the experiment. This dataframe was derived from "spill.char.dataset2.csv" through the R code "PredictingPersistenceMechModel.R". Supplemental Figure 1. The dataset "spill.char.dataset1.csv" contains the data depicted in Figure S1. This figure plots, by strain, the duration of virus detection in the passage experiment "trans.ability" against maximum likelihood infection prevalence on the spillover plate "best.p". The code to make this graph is contained in "SpilloverCharacteristicsDataPrep.R" L377-382. Supplemental Figure 2. The dataset "spill.char.dataset1.csv" contains the data depicted in Figure S2. This figure plots, by strain, the duration of virus detection in the passage experiment "trans.ability" against infection intensity of infected worms on the spillover plate "median.ct". The code to make this graph is contained in "SpilloverCharacteristicsDataPrep.R" L391-396. Supplemental Figure 3. The dataset "spill.char.dataset1.csv" contains the data depicted in Figure S3. This figure plots, by strain, the duration of virus detection in the passage experiment "trans.ability" against maximum likelihood shedding ability of worms on the spillover plate "best.y". The code to make this graph is contained in "SpilloverCharacteristicsDataPrep.R" L384-389. Supplemental Figure 4. This figure displays correlations among the 4 epidemiological characteristics of spillover that were studied. The dataset "spill.char.dataset1.csv" contains the data depicted in Figure S4. The code to make this graph is contained in "PredictingPersistenceCorrelativeModels.R" L160-179. Supplemental Figure 5. This figure displays the changes to the infection intensity estimate when we corrected for the possibility of multiple infected worms in group extractions. In Figure S5A, the corrected infection intensity is plotted against the uncorrected infection intensity, and in figure S5B, corrected and uncorrect infection intensities are plotted against infection prevalence. The code for producing this figure is found in "SpilloverCharacteristicsDataPrep.R" L258-313. Supplemental Figure 6. This figure displays maximum likelihood median tissue culture infectious dose (TCID50) estimates for each strain and thaw line used in the experiment. The data is contained in "TCID50bythaw.csv", and code to produce this figure can be found in "SpilloverCharacteristicsDataPrep.R" L403-440. Supplemental Figure 7. This figure displays the predicted median qPCR Ct values for infected worms (predicted with a linear mixed effects model with maximum likelihood prevalence as a fixed effect and strain as a random effect) plotted against the actual corrected median ct values. The linear mixed effects model was used to ascertain the standard deviation of infection intensity used in our mechanistic model. The data is contained in "spill.char.dataset1.csv", and code to produce this figure is found in "SpilloverCharacteristics.DataPrep.R" L327-368.