In order to identify factors affecting bulk fitness of the influenza A virus mutants, we performed multiple linear regression analyses. The viral fitness determined in the EMPIRIC experiments is correlated with _in vitro_ protein abundance (R² = 0.53, p < 10−2). To look if additional factors explain the observed variance of the fitness, we use a linear regression model using protein abundance, ELISA assay result, and mRNA folding energy dG as explanatory variables. The combination of protein abundance and ELISA data improved the prediction of fitness, R² = 0.68, ANOVA comparison of the models yields p = 0.014. The best prediction was achieved by a linear model combining all three factors, protein abundance, ELISA, and mRNA dG, yielding R² = 0.752, and marginally significant p = 0.053 for ANOVA comparison of the 2 vs. 3 factor models. The folding energy of mRNA was defined as the free energy dG of the viral mRNA segment consisting of nucleotides -4 to 37 relative to the HA start codon, as described in . Adding a categorical variable representing DNA library comprising a particular mutant did not improve the fit, suggesting the measurements were not affected by batch effect. Likewise, inclusion of additional metrics such as HA vRNA abundance normalized to NA viral RNA, vRNA dG, position of the mutation, etc., did not improve the predictive power of the multiple linear regression model for fitness (data not shown). These results suggest that the fitness of synonymous mutants of IAV is affected by multiple factors, likely having a complex relationship with RNA structure and functional interactions between the viral and host components.

MAIN TEXT CONTRIBUTION [FROM ANETH MANUSCRIPT START] In order to further investigate potential differences in the fitness effects of silent mutations, we performed infection kinetics experiments on 4 silent substitutions, L5LCTC, L6LCTC, L6LTTG and L8LCTA. Of note, three of these mutants exhibited growth defects in bulk competition (L6LCTC,L6LTTG and L8LCTA),while L5LCTC did not exhibit a large growth effect in the bulk competition (Figure 2A). [FROM ANETH MANUSCRIPT OVER] We used phenomenological model of viral kinetics to fit our experimental data in order to quantitatively characterize infection kinetics for the mutants (see Methods/Supplement for details). Sensitivity analysis of the model (see Supplement) showed that we can reliably estimate only 2 model parameters (out of 5): β rate of infection of target cells and τE average duration of an eclipse phase (time infected cells spend before they start yielding virus), the results are presented in a Table [table:kinetics]. According to the model fit, L6LTTG demonstrates WT-like kinetics of viral yield, while other 3 mutants L5LCTC, L6LCTC, and L8LCTA demonstrate delayed eclipse phase along with improved infection rate, while being virtually indistinguishable from each other. [FROM ANETH MANUSCRIPT START] The discrepancy between the viral kinetic assay and bulk competitions for the L5LCTC mutation could be due to inherent challenges in measuring fitness effects with precision using these techniques. This is consistent with our analyses of the reproducibility of estimates of fitness effects that indicate that the precision of these estimates can distinguish overall trends, but are not precise enough to discern the fitness effect of a precise mutation with strong confidence. For this reason, we have taken multiple approaches to investigate fitness effects. The slowed infection kinetics of this panel of four individual mutations lends additional evidence that silent mutations in the signal sequence can cause strong fitness defects. [FROM ANETH MANUSCRIPT OVER]