Immune correlates of protection primary analysis.

Associations between immune responses and SHIV infection risk were assessed by Cox proportional hazards models for assays that remained after filtering out highly correlated assays. Results are grouped by related immunologic assays and time point, represented by different tabs within the spreadsheet. Sex and vaccine arm were included as covariates, along with a sex by assay interaction term to allow for differing patterns of protection between sexes. Assays with protective effects that met the significance threshold of unadjusted P<0.05 and FDR-adjusted q<0.2 are highlighted in blue, while assays with P<0.1 but q>0.2 are highlighted in light purple. Analyses that included active arms only (“vaccinees only”), were baseline-controlled, and adjusted for outliers are indicated. The proportional hazards (PH) assumption was tested for the sex-specific and combined-sex models; small p-values indicate that this assumption may be invalid, which would prompt additional caution in interpreting the results, particularly the p-values. Abbreviations: SD, standard deviation; HR, hazard ratio. (XLSX)