Additional file 1 of Modelling upper respiratory viral load dynamics of SARS-CoV-2

Additional file 1: Supplementary Table S1. Summary of studies which were identified during the literature search but did not provide viral load data when the corresponding authors were contacted. Studies included both male and female patients, unless stated. All ages stated are in years. *For some studies, particularly those carried out early in the pandemic, patients were hospitalised to ensure isolation (i.e. cohorts may include asymptomatic subjects or those with very mild symptoms). ��� These studies were identified during our literature search, but we were informed by the study authors that quantitative viral load was not available (i.e. only positive or negative result recorded). Supplementary Table S2. Summary of antiviral and immunomodulatory treatment in the studies included in analysis. Supplementary Table S3. Summary of the population-level (i.e. not study- or patient-specific) parameter values (and 95% credible intervals) obtained for the multi-level regression modelling (as displayed in Fig. 2). Patient- and study-specific random effects were used for both the peak (log-transformed) viral load, and its rate of decline per day. Supplementary Table S4. assessing the goodness of fit of the regression models using leave-one-out cross-validation. Supplementary Table S5. Summary of the population-level (i.e. not study- or patient-specific) parameter values (and 95% credible intervals) obtained for the mechanistic viral load model (Eqs. 6���8). Samples from k a 0 $$ {k}_a^0 $$ and I max 0 $$ {I}_{\mathrm{max}}^0 $$ were used to generate the black line and dark grey shaded area in Fig. 4. Supplementary Figure S1. Standard curves relating cycle-threshold (Ct) values to viral load. Seven standard curves, identified from published studies (see Methods) are plotted. Supplementary Figure S2. Summary of all the data collected (see Table 1 in the main text). For the studies shown in blue, viral loads have been estimated using an averaged standard curve (see Methods for details). Supplementary Figure S3 Comparison of timing of first sample and viral load by severity. Supplementary Figure S4. Estimations of the statistical power in the regression analyses. Supplementary Figure S5. Relationship between patient-specific parameters governing the immune response in the mechanistic model and disease severity. Supplementary Figure S6. Posterior means and 95% credible intervals for the study-specific offsets in the mechanistic model. Supplementary Figure S7. A comparison of the prior and posterior distributions for the early and late immune responses in the mechanistic model. Supplementary Figure S8. (shown over the following 7 pages): Output from the mechanistic model alongside the data, for all 155 patients considered. In the heading of each panel, the first number indicates the study (studies numbered as in Table 1).