Fig 1.
Susceptibility to ISP across a panel of 64 Staphylococcaceae isolates, measured using three distinct methods.
Bar lengths and colour show the mean change in ISP assessed by plaque assay (PFU/μL), optical density (proportion decrease in OD with infection after 24 hours), and qPCR (log10 fold change in viral load after 24 hours), with error bars representing the standard error of the mean across at least four biological replicates. The phylogeny of Staphylococcaceae hosts is presented on the left, with the scale bar representing the number of nucleotide substitutions per site. Strain names are presented on the right, with non-aureus species in bold and the propagation host labelled with an asterisk (a full version of the tree is available at https://doi.org/10.6084/m9.figshare.21642209.v1).
Fig 2.
Ancestral state reconstruction of susceptibility to ISP measured by OD.
Ancestral states were estimated from model (2) for each node and plotted in colour across the Staphylococcaceae host phylogeny, with the scale bar representing nucleotide substitutions per site. Colours represent susceptibility of a host to infection with ISP measured by OD (proportion change in optical density in infected compared to non-infected cultures), with black representing the lowest level of susceptibility and yellow the highest. Strain IDs are presented on the right, with non-aureus species in bold and the propagation host labelled with an asterisk (*).
Fig 3.
Ancestral state reconstruction of the S. aureus host strains susceptibility to ISP measured by OD.
Ancestral states were estimated from model (2) for each node and plotted in colour across the Staphylococcus aureus host phylogeny, with the scale bar representing nucleotide substitutions per site. Colours represent susceptibility of a host to infection with ISP measured by OD (proportion change in optical density in infected compared to non-infected cultures), with black representing the lowest level of susceptibility and yellow the highest. Strain IDs are presented on the right and the propagation host labelled with an asterisk (*).
Table 1.
Estimates for the repeatability and phylogenetic heritability across 64 Staphylococcaceae isolates.
Estimates of repeatability are taken from model (2) and estimates of phylogenetic heritability (the proportion of phylogenetic and non-phylogenetic strain variation explained by the host phylogeny) and phylogenetic heritability of the total variance (the proportion of total variation explained by the host phylogeny) are taken from model (1). PA = plaque assay, CI = credible interval.
Table 2.
Inter-strain correlations in susceptibility measures between methods.
Numbers show the mean estimates for the correlation strength (r, white cells) and slope (β, grey cells) between pairs of methods, with 95% credible intervals (CIs) indicated in brackets. The slopes were calculated with the variables in columns as x and variables in rows as y. Estimates with CIs that do not span zero are highlighted in bold. PA = plaque assay, *value on a probit scale.
Fig 4.
Leave-one-out cross-validation of phylogenetic mixed models fitted to plaque assay, OD, and qPCR data.
Predictions of “unknown” trait values from a leave-one-out cross-validation of phylogenetic model (2) for binary plaque assay (A), continuous (PFU/μL) plaque assay (B), change in OD with infection (C) and fold changes in viral load measured by qPCR (D). Each datapoint represents an individual strain of Staphylococcaceae whose measured trait value has been removed from model (2) and predicted from its evolutionary relationships to other Staphylococcaceae isolates and their trait values. Solid diagonal lines illustrate the location of 1:1 predictions and dotted lines indicate the root-mean-squared errors around these lines. Orange vertical lines represent the predicted trait values of all strains from a null (intercept only) model. For panels B, C, and D, points plotted in pink show the S. aureus strains whereas points plotted in grey show the non-S. aureus strains.