Skip to main content
. Author manuscript; available in PMC: 2014 Mar 21.
Published in final edited form as: J Stat Mech. 2013 Jan 16;2013:P01010. doi: 10.1088/1742-5468/2013/01/P01010

Figure 3.

Figure 3

Impact of constraining the initial mutant frequency, f0, on the kinetics of viral escape. In panel A we show fits of the model (eqn. (10)) to the sequence data obtained assuming that the initial frequency of escape variant is lower than fc = 10−4. The fit is done by adding an extra penalizing term 10105(f0−fc) to the log-likelihood (eqn. (16)). The constraint leads to much higher estimates of the escape rate (Table 3) but in a poor description of the data. In panel B in addition to the constraint to the initial mutant frequency we allow the escape rate to decline over the course of infection (eqn. (13)). This extension improves the fit of the constrained model to data for 3 out 5 epitopes (Table 3).