Using heterogeneity in the population structure of U.S. swine farms to compare transmission models for porcine epidemic diarrhoea

Eamon B. O’Dea; Harry Snelson; Shweta Bansal

. CC-BY 4.0 International license is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. It . https://doi.org/10.1101/017178 doi: bioRxiv preprint between the lag-1 cross correlation matrix and the distance, shared border, and transport flow matrices. Figure 2 shows that parameters with the highest sensitivities were the transmission probability for transport edges, the amplitude of seasonal variation in transmission probabilities, and the resolution of the raster grid used to generate spatial edges. All correlation types typically increased with transmission probabilities across transport edges, with those of the transport flow matrix increasing the most (Supplementary Fig. S3 ). All correlation types typically increased as spatial resolution decreased and the spatial contact network became more dense (Supplementary Fig. S4 ). Sensitivity to the spatial transmission probability was noticeable only on the more dense spatial network (Supplementary Fig. S5 ). Seasonal variation in the transmission probabilities tended to reduce correlations (Supplementary Figs. S3 and S4 ). Correlation size was indicative of statistical significance ( Supplementary Figs. S3-S5 ). In summary, the largest sensitivities were to transmission and contact parameters, and these parameters had unsurprising relationships with the matrix correlations.