Inferring pandemic growth rates from sequence data

de Silva, Eric; Ferguson, Neil M.; Fraser, Christophe

Our results offer some support for the hypothesis that the inferred slowing of growth in the effective population size of H1N1 seen in the USA during late April/early May 2009 revealed by the BSP shown in figure 1 is not an artefact. However, given the very real impact sampling epidemiologically linked sequences has on inferred growth rates (as highlighted in our non-random sampling of simulated sequences), increasing epidemiological linkage in samples being collected cannot be ruled out as a contributory cause of the inferred slowing of epidemic growth. Future use of these methods would therefore be considerably aided by more detailed contextual epidemiological data on individuals from whom virological samples are collected and sequenced.